Docs: How to use MinerU to parse pdf documents (#10763)

### What problem does this PR solve?



### Type of change

- [x] Documentation Update

.github/workflows/release.yml

@@ -25,7 +25,7 @@ jobs:
       - name: Check out code
         uses: actions/checkout@v4
         with:
-          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          token: ${{ secrets.GITHUB_TOKEN }}  # Use the secret as an environment variable
           fetch-depth: 0
           fetch-tags: true
 
@@ -69,7 +69,7 @@ jobs:
         # https://github.com/actions/upload-release-asset has been replaced by https://github.com/softprops/action-gh-release
         uses: softprops/action-gh-release@v2
         with:
-          token: ${{ secrets.MY_GITHUB_TOKEN }}  # Use the secret as an environment variable
+          token: ${{ secrets.GITHUB_TOKEN }}  # Use the secret as an environment variable
           prerelease: ${{ env.PRERELEASE }}
           tag_name: ${{ env.RELEASE_TAG }}
           # The body field does not support environment variable substitution directly.
@@ -120,3 +120,17 @@ jobs:
           packages-dir: sdk/python/dist/
           password: ${{ secrets.PYPI_API_TOKEN }}
           verbose: true
+
+      - name: Build ragflow-cli
+        if: startsWith(github.ref, 'refs/tags/v')
+        run: |
+          cd admin/client && \
+          uv build
+
+      - name: Publish client package distributions to PyPI
+        if: startsWith(github.ref, 'refs/tags/v')
+        uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          packages-dir: admin/client/dist/
+          password: ${{ secrets.PYPI_API_TOKEN }}
+          verbose: true

.github/workflows/tests.yml

@@ -34,12 +34,10 @@ jobs:
       # https://github.com/hmarr/debug-action
       #- uses: hmarr/debug-action@v2
 
-      - name: Show who triggered this workflow
+      - name: Ensure workspace ownership
         run: |
           echo "Workflow triggered by ${{ github.event_name }}"
-
+          echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
-      - name: Ensure workspace ownership
-        run: echo "chown -R $USER $GITHUB_WORKSPACE" && sudo chown -R $USER $GITHUB_WORKSPACE
 
       # https://github.com/actions/checkout/issues/1781
       - name: Check out code
@@ -48,6 +46,44 @@ jobs:
           fetch-depth: 0
           fetch-tags: true
 
+      - name: Check workflow duplication
+        if: ${{ !cancelled() && !failure() && (github.event_name != 'pull_request' || contains(github.event.pull_request.labels.*.name, 'ci')) }}
+        run: |
+          if [[ ${{ github.event_name }} != 'pull_request' ]]; then
+            HEAD=$(git rev-parse HEAD)
+            # Find a PR that introduced a given commit
+            gh auth login --with-token <<< "${{ secrets.GITHUB_TOKEN }}"
+            PR_NUMBER=$(gh pr list --search ${HEAD} --state merged --json number --jq .[0].number)
+            echo "HEAD=${HEAD}"
+            echo "PR_NUMBER=${PR_NUMBER}"
+            if [[ -n ${PR_NUMBER} ]]; then
+              PR_SHA_FP=${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}/PR_${PR_NUMBER}
+              if [[ -f ${PR_SHA_FP} ]]; then
+                read -r PR_SHA PR_RUN_ID < "${PR_SHA_FP}"
+                # Calculate the hash of the current workspace content
+                HEAD_SHA=$(git rev-parse HEAD^{tree})
+                if [[ ${HEAD_SHA} == ${PR_SHA} ]]; then
+                  echo "Cancel myself since the workspace content hash is the same with PR #${PR_NUMBER} merged. See ${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}/actions/runs/${PR_RUN_ID} for details."
+                  gh run cancel ${GITHUB_RUN_ID}
+                  while true; do
+                    status=$(gh run view ${GITHUB_RUN_ID} --json status -q .status)
+                    [ "$status" = "completed" ] && break
+                    sleep 5
+                  done
+                  exit 1
+                fi
+              fi
+            fi
+          else
+            PR_NUMBER=${{ github.event.pull_request.number }}
+            PR_SHA_FP=${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}/PR_${PR_NUMBER}
+            # Calculate the hash of the current workspace content
+            PR_SHA=$(git rev-parse HEAD^{tree})
+            echo "PR #${PR_NUMBER} workspace content hash: ${PR_SHA}"
+            mkdir -p ${RUNNER_WORKSPACE_PREFIX}/artifacts/${GITHUB_REPOSITORY}
+            echo "${PR_SHA} ${GITHUB_RUN_ID}" > ${PR_SHA_FP}
+          fi
+
       # https://github.com/astral-sh/ruff-action
       - name: Static check with Ruff
         uses: astral-sh/ruff-action@v3
@@ -59,11 +95,11 @@ jobs:
         run: |
           RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-$HOME}
           sudo docker pull ubuntu:22.04
-          sudo docker build --progress=plain --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
+          sudo DOCKER_BUILDKIT=1 docker build --build-arg LIGHTEN=1 --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly-slim .
 
       - name: Build ragflow:nightly
         run: |
-          sudo docker build --progress=plain --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
+          sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 -f Dockerfile -t infiniflow/ragflow:nightly .
 
       - name: Start ragflow:nightly-slim
         run: |

.gitignore

@@ -149,7 +149,7 @@ out
 # Nuxt.js build / generate output
 .nuxt
 dist
-
+ragflow_cli.egg-info
 # Gatsby files
 .cache/
 # Comment in the public line in if your project uses Gatsby and not Next.js

Dockerfile

@@ -191,6 +191,7 @@ ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"
 ENV PYTHONPATH=/ragflow/
 
 COPY web web
+COPY admin admin
 COPY api api
 COPY conf conf
 COPY deepdoc deepdoc

README.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -84,8 +84,8 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Latest Updates
 
+- 2025-10-15 Supports orchestrable ingestion pipeline.
 - 2025-08-08 Supports OpenAI's latest GPT-5 series models.
-- 2025-08-04 Supports new models, including Kimi K2 and Grok 4.
 - 2025-08-01 Supports agentic workflow and MCP.
 - 2025-05-23 Adds a Python/JavaScript code executor component to Agent.
 - 2025-05-05 Supports cross-language query.
@@ -135,7 +135,7 @@ releases! 🌟
 ## 🔎 System Architecture
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Get Started
@@ -187,7 +187,7 @@ releases! 🌟
 > All Docker images are built for x86 platforms. We don't currently offer Docker images for ARM64.
 > If you are on an ARM64 platform, follow [this guide](https://ragflow.io/docs/dev/build_docker_image) to build a Docker image compatible with your system.
 
-   > The command below downloads the `v0.20.5-slim` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.20.5-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server. For example: set `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` for the full edition `v0.20.5`.
+   > The command below downloads the `v0.21.1-slim` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.21.1-slim`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server. For example: set `RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1` for the full edition `v0.21.1`.
 
    ```bash
    $ cd ragflow/docker
@@ -200,8 +200,8 @@ releases! 🌟
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    |-------------------|-----------------|-----------------------|--------------------------|
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                   | Stable release            |
+   | v0.21.1-slim      | &approx;2       | ❌                   | Stable release            |
    | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
    | nightly-slim      | &approx;2       | ❌                   | _Unstable_ nightly build  |
 
@@ -341,11 +341,13 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 5. If your operating system does not have jemalloc, please install it as follows:
 
    ```bash
-   # ubuntu
+   # Ubuntu
    sudo apt-get install libjemalloc-dev
-   # centos
+   # CentOS
    sudo yum install jemalloc
-   # mac
+   # OpenSUSE
+   sudo zypper install jemalloc
+   # macOS
    sudo brew install jemalloc
    ```
 

README_id.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="Logo ragflow">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="Logo ragflow">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Lencana Daring" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Rilis%20Terbaru" alt="Rilis Terbaru">
@@ -80,8 +80,8 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Pembaruan Terbaru
 
+- 2025-10-15 Dukungan untuk jalur data yang terorkestrasi.
 - 2025-08-08 Mendukung model seri GPT-5 terbaru dari OpenAI.
-- 2025-08-04 Mendukung model baru, termasuk Kimi K2 dan Grok 4.
 - 2025-08-01 Mendukung alur kerja agen dan MCP.
 - 2025-05-23 Menambahkan komponen pelaksana kode Python/JS ke Agen.
 - 2025-05-05 Mendukung kueri lintas bahasa.
@@ -129,7 +129,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 ## 🔎 Arsitektur Sistem
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Mulai
@@ -181,7 +181,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 > Semua gambar Docker dibangun untuk platform x86. Saat ini, kami tidak menawarkan gambar Docker untuk ARM64.
 > Jika Anda menggunakan platform ARM64, [silakan gunakan panduan ini untuk membangun gambar Docker yang kompatibel dengan sistem Anda](https://ragflow.io/docs/dev/build_docker_image).
 
-> Perintah di bawah ini mengunduh edisi v0.20.5-slim dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.20.5-slim, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server. Misalnya, atur RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5 untuk edisi lengkap v0.20.5.
+> Perintah di bawah ini mengunduh edisi v0.21.1-slim dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.21.1-slim, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server. Misalnya, atur RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1 untuk edisi lengkap v0.21.1.
 
 ```bash
 $ cd ragflow/docker
@@ -194,8 +194,8 @@ $ docker compose -f docker-compose.yml up -d
 
 | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
 | ----------------- | --------------- | --------------------- | ------------------------ |
-| v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+| v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-| v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+| v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
 | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
 | nightly-slim      | &approx;2       | ❌                    | _Unstable_ nightly build |
 

README_ja.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -60,8 +60,8 @@
 
 ## 🔥 最新情報
 
+- 2025-10-15 オーケストレーションされたデータパイプラインのサポート。
 - 2025-08-08 OpenAI の最新 GPT-5 シリーズモデルをサポートします。
-- 2025-08-04 新モデル、キミK2およびGrok 4をサポート。
 - 2025-08-01 エージェントワークフローとMCPをサポート。
 - 2025-05-23 エージェントに Python/JS コードエグゼキュータコンポーネントを追加しました。
 - 2025-05-05 言語間クエリをサポートしました。
@@ -109,7 +109,7 @@
 ## 🔎 システム構成
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 初期設定
@@ -160,7 +160,7 @@
 > 現在、公式に提供されているすべての Docker イメージは x86 アーキテクチャ向けにビルドされており、ARM64 用の Docker イメージは提供されていません。
 > ARM64 アーキテクチャのオペレーティングシステムを使用している場合は、[このドキュメント](https://ragflow.io/docs/dev/build_docker_image)を参照して Docker イメージを自分でビルドしてください。
 
-   > 以下のコマンドは、RAGFlow Docker イメージの v0.20.5-slim エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.20.5-slim とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。例えば、完全版 v0.20.5 をダウンロードするには、RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5 と設定します。
+   > 以下のコマンドは、RAGFlow Docker イメージの v0.21.1-slim エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.21.1-slim とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。例えば、完全版 v0.21.1 をダウンロードするには、RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1 と設定します。
 
    ```bash
    $ cd ragflow/docker
@@ -173,8 +173,8 @@
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
    | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
    | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
 

README_ko.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -60,8 +60,8 @@
 
 ## 🔥 업데이트
 
+- 2025-10-15 조정된 데이터 파이프라인 지원.
 - 2025-08-08 OpenAI의 최신 GPT-5 시리즈 모델을 지원합니다.
-- 2025-08-04 새로운 모델인 Kimi K2와 Grok 4를 포함하여 지원합니다.
 - 2025-08-01 에이전트 워크플로우와 MCP를 지원합니다.
 - 2025-05-23 Agent에 Python/JS 코드 실행기 구성 요소를 추가합니다.
 - 2025-05-05 언어 간 쿼리를 지원합니다.
@@ -109,7 +109,7 @@
 ## 🔎 시스템 아키텍처
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 시작하기
@@ -160,7 +160,7 @@
 > 모든 Docker 이미지는 x86 플랫폼을 위해 빌드되었습니다. 우리는 현재 ARM64 플랫폼을 위한 Docker 이미지를 제공하지 않습니다.
 > ARM64 플랫폼을 사용 중이라면, [시스템과 호환되는 Docker 이미지를 빌드하려면 이 가이드를 사용해 주세요](https://ragflow.io/docs/dev/build_docker_image).
 
-   > 아래 명령어는 RAGFlow Docker 이미지의 v0.20.5-slim 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.20.5-slim과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오. 예를 들어, 전체 버전인 v0.20.5을 다운로드하려면 RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5로 설정합니다.
+   > 아래 명령어는 RAGFlow Docker 이미지의 v0.21.1-slim 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.21.1-slim과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오. 예를 들어, 전체 버전인 v0.21.1을 다운로드하려면 RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1로 설정합니다.
 
    ```bash
    $ cd ragflow/docker
@@ -173,8 +173,8 @@
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
    | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
    | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
 

README_pt_br.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="520" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Badge Estático" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Última%20Relese" alt="Última Versão">
@@ -80,8 +80,8 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Últimas Atualizações
 
+- 10-15-2025 Suporte para pipelines de dados orquestrados.
 - 08-08-2025 Suporta a mais recente série GPT-5 da OpenAI.
-- 04-08-2025 Suporta novos modelos, incluindo Kimi K2 e Grok 4.
 - 01-08-2025 Suporta fluxo de trabalho agente e MCP.
 - 23-05-2025 Adicione o componente executor de código Python/JS ao Agente.
 - 05-05-2025 Suporte a consultas entre idiomas.
@@ -129,7 +129,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 ## 🔎 Arquitetura do Sistema
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 Primeiros Passos
@@ -180,7 +180,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 > Todas as imagens Docker são construídas para plataformas x86. Atualmente, não oferecemos imagens Docker para ARM64.
 > Se você estiver usando uma plataforma ARM64, por favor, utilize [este guia](https://ragflow.io/docs/dev/build_docker_image) para construir uma imagem Docker compatível com o seu sistema.
 
-    > O comando abaixo baixa a edição `v0.20.5-slim` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.20.5-slim`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor. Por exemplo: defina `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` para a edição completa `v0.20.5`.
+    > O comando abaixo baixa a edição `v0.21.1-slim` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.21.1-slim`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor. Por exemplo: defina `RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1` para a edição completa `v0.21.1`.
 
     ```bash
     $ cd ragflow/docker
@@ -193,8 +193,8 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
     | Tag da imagem RAGFlow | Tamanho da imagem (GB) | Possui modelos de incorporação? | Estável?                 |
     | --------------------- | ---------------------- | ------------------------------- | ------------------------ |
-    | v0.20.5               | ~9                     | :heavy_check_mark:              | Lançamento estável       |
+    | v0.21.1               | ~9                     | :heavy_check_mark:              | Lançamento estável       |
-    | v0.20.5-slim          | ~2                     | ❌                              | Lançamento estável       |
+    | v0.21.1-slim          | ~2                     | ❌                              | Lançamento estável       |
     | nightly               | ~9                     | :heavy_check_mark:              | _Instável_ build noturno |
     | nightly-slim          | ~2                     | ❌                               | _Instável_ build noturno |
 

README_tzh.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -83,8 +83,8 @@
 
 ## 🔥 近期更新
 
+- 2025-10-15 支援可編排的資料管道。
 - 2025-08-08 支援 OpenAI 最新的 GPT-5 系列模型。
-- 2025-08-04 支援 Kimi K2 和 Grok 4 等模型.
 - 2025-08-01 支援 agentic workflow 和 MCP
 - 2025-05-23 為 Agent 新增 Python/JS 程式碼執行器元件。
 - 2025-05-05 支援跨語言查詢。
@@ -132,7 +132,7 @@
 ## 🔎 系統架構
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 快速開始
@@ -183,7 +183,7 @@
 > 所有 Docker 映像檔都是為 x86 平台建置的。目前，我們不提供 ARM64 平台的 Docker 映像檔。
 > 如果您使用的是 ARM64 平台，請使用 [這份指南](https://ragflow.io/docs/dev/build_docker_image) 來建置適合您系統的 Docker 映像檔。
 
-   > 執行以下指令會自動下載 RAGFlow slim Docker 映像 `v0.20.5-slim`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.20.5-slim` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。例如，你可以透過設定 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` 來下載 RAGFlow 鏡像的 `v0.20.5` 完整發行版。
+   > 執行以下指令會自動下載 RAGFlow slim Docker 映像 `v0.21.1-slim`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.21.1-slim` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。例如，你可以透過設定 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1` 來下載 RAGFlow 鏡像的 `v0.21.1` 完整發行版。
 
    ```bash
    $ cd ragflow/docker
@@ -196,8 +196,8 @@
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
    | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
    | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
 

README_zh.md

@@ -1,6 +1,6 @@
 <div align="center">
 <a href="https://demo.ragflow.io/">
-<img src="web/src/assets/logo-with-text.png" width="350" alt="ragflow logo">
+<img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
 
@@ -22,7 +22,7 @@
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.20.5">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.21.1">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -83,8 +83,8 @@
 
 ## 🔥 近期更新
 
-- 2025-08-08 支持 OpenAI 最新的 GPT-5 系列模型.
+- 2025-10-15 支持可编排的数据管道。
-- 2025-08-04 新增对 Kimi K2 和 Grok 4 等模型的支持.
+- 2025-08-08 支持 OpenAI 最新的 GPT-5 系列模型。
 - 2025-08-01 支持 agentic workflow 和 MCP。
 - 2025-05-23 Agent 新增 Python/JS 代码执行器组件。
 - 2025-05-05 支持跨语言查询。
@@ -132,7 +132,7 @@
 ## 🔎 系统架构
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
-<img src="https://github.com/infiniflow/ragflow/assets/12318111/d6ac5664-c237-4200-a7c2-a4a00691b485" width="1000"/>
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
 </div>
 
 ## 🎬 快速开始
@@ -183,7 +183,7 @@
 > 请注意，目前官方提供的所有 Docker 镜像均基于 x86 架构构建，并不提供基于 ARM64 的 Docker 镜像。
 > 如果你的操作系统是 ARM64 架构，请参考[这篇文档](https://ragflow.io/docs/dev/build_docker_image)自行构建 Docker 镜像。
 
-   > 运行以下命令会自动下载 RAGFlow slim Docker 镜像 `v0.20.5-slim`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.20.5-slim` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。比如，你可以通过设置 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.20.5` 来下载 RAGFlow 镜像的 `v0.20.5` 完整发行版。
+   > 运行以下命令会自动下载 RAGFlow slim Docker 镜像 `v0.21.1-slim`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.21.1-slim` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。比如，你可以通过设置 `RAGFLOW_IMAGE=infiniflow/ragflow:v0.21.1` 来下载 RAGFlow 镜像的 `v0.21.1` 完整发行版。
 
    ```bash
    $ cd ragflow/docker
@@ -196,8 +196,8 @@
 
    | RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?                  |
    | ----------------- | --------------- | --------------------- | ------------------------ |
-   | v0.20.5           | &approx;9       | :heavy_check_mark:    | Stable release           |
+   | v0.21.1           | &approx;9       | :heavy_check_mark:    | Stable release           |
-   | v0.20.5-slim      | &approx;2       | ❌                    | Stable release           |
+   | v0.21.1-slim      | &approx;2       | ❌                    | Stable release           |
    | nightly           | &approx;9       | :heavy_check_mark:    | _Unstable_ nightly build |
    | nightly-slim      | &approx;2       | ❌                     | _Unstable_ nightly build |
 

admin/admin_client.py

@@ -1,567 +0,0 @@
-import argparse
-import base64
-from Cryptodome.PublicKey import RSA
-from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
-from typing import Dict, List, Any
-from lark import Lark, Transformer, Tree
-import requests
-from requests.auth import HTTPBasicAuth
-
-GRAMMAR = r"""
-start: command
-
-command: sql_command | meta_command
-
-sql_command: list_services
-           | show_service
-           | startup_service
-           | shutdown_service
-           | restart_service
-           | list_users
-           | show_user
-           | drop_user
-           | alter_user
-           | create_user
-           | activate_user
-           | list_datasets
-           | list_agents
-
-// meta command definition
-meta_command: "\\" meta_command_name [meta_args]
-
-meta_command_name: /[a-zA-Z?]+/
-meta_args: (meta_arg)+
-
-meta_arg: /[^\\s"']+/ | quoted_string
-
-// command definition
-
-LIST: "LIST"i
-SERVICES: "SERVICES"i
-SHOW: "SHOW"i
-CREATE: "CREATE"i
-SERVICE: "SERVICE"i
-SHUTDOWN: "SHUTDOWN"i
-STARTUP: "STARTUP"i
-RESTART: "RESTART"i
-USERS: "USERS"i
-DROP: "DROP"i
-USER: "USER"i
-ALTER: "ALTER"i
-ACTIVE: "ACTIVE"i
-PASSWORD: "PASSWORD"i
-DATASETS: "DATASETS"i
-OF: "OF"i
-AGENTS: "AGENTS"i
-
-list_services: LIST SERVICES ";"
-show_service: SHOW SERVICE NUMBER ";"
-startup_service: STARTUP SERVICE NUMBER ";"
-shutdown_service: SHUTDOWN SERVICE NUMBER ";"
-restart_service: RESTART SERVICE NUMBER ";"
-
-list_users: LIST USERS ";"
-drop_user: DROP USER quoted_string ";"
-alter_user: ALTER USER PASSWORD quoted_string quoted_string ";"
-show_user: SHOW USER quoted_string ";"
-create_user: CREATE USER quoted_string quoted_string ";"
-activate_user: ALTER USER ACTIVE quoted_string status ";"
-
-list_datasets: LIST DATASETS OF quoted_string ";"
-list_agents: LIST AGENTS OF quoted_string ";"
-
-identifier: WORD
-quoted_string: QUOTED_STRING
-status: WORD
-
-QUOTED_STRING: /'[^']+'/ | /"[^"]+"/
-WORD: /[a-zA-Z0-9_\-\.]+/
-NUMBER: /[0-9]+/
-
-%import common.WS
-%ignore WS
-"""
-
-
-class AdminTransformer(Transformer):
-
-    def start(self, items):
-        return items[0]
-
-    def command(self, items):
-        return items[0]
-
-    def list_services(self, items):
-        result = {'type': 'list_services'}
-        return result
-
-    def show_service(self, items):
-        service_id = int(items[2])
-        return {"type": "show_service", "number": service_id}
-
-    def startup_service(self, items):
-        service_id = int(items[2])
-        return {"type": "startup_service", "number": service_id}
-
-    def shutdown_service(self, items):
-        service_id = int(items[2])
-        return {"type": "shutdown_service", "number": service_id}
-
-    def restart_service(self, items):
-        service_id = int(items[2])
-        return {"type": "restart_service", "number": service_id}
-
-    def list_users(self, items):
-        return {"type": "list_users"}
-
-    def show_user(self, items):
-        user_name = items[2]
-        return {"type": "show_user", "username": user_name}
-
-    def drop_user(self, items):
-        user_name = items[2]
-        return {"type": "drop_user", "username": user_name}
-
-    def alter_user(self, items):
-        user_name = items[3]
-        new_password = items[4]
-        return {"type": "alter_user", "username": user_name, "password": new_password}
-
-    def create_user(self, items):
-        user_name = items[2]
-        password = items[3]
-        return {"type": "create_user", "username": user_name, "password": password, "role": "user"}
-
-    def activate_user(self, items):
-        user_name = items[3]
-        activate_status = items[4]
-        return {"type": "activate_user", "activate_status": activate_status, "username": user_name}
-
-    def list_datasets(self, items):
-        user_name = items[3]
-        return {"type": "list_datasets", "username": user_name}
-
-    def list_agents(self, items):
-        user_name = items[3]
-        return {"type": "list_agents", "username": user_name}
-
-    def meta_command(self, items):
-        command_name = str(items[0]).lower()
-        args = items[1:] if len(items) > 1 else []
-
-        # handle quoted parameter
-        parsed_args = []
-        for arg in args:
-            if hasattr(arg, 'value'):
-                parsed_args.append(arg.value)
-            else:
-                parsed_args.append(str(arg))
-
-        return {'type': 'meta', 'command': command_name, 'args': parsed_args}
-
-    def meta_command_name(self, items):
-        return items[0]
-
-    def meta_args(self, items):
-        return items
-
-
-def encode_to_base64(input_string):
-    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
-    return base64_encoded.decode('utf-8')
-
-
-def encrypt(input_string):
-    pub = '-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----'
-    pub_key = RSA.importKey(pub)
-    cipher = Cipher_pkcs1_v1_5.new(pub_key)
-    cipher_text = cipher.encrypt(base64.b64encode(input_string.encode('utf-8')))
-    return base64.b64encode(cipher_text).decode("utf-8")
-
-
-class AdminCommandParser:
-    def __init__(self):
-        self.parser = Lark(GRAMMAR, start='start', parser='lalr', transformer=AdminTransformer())
-        self.command_history = []
-
-    def parse_command(self, command_str: str) -> Dict[str, Any]:
-        if not command_str.strip():
-            return {'type': 'empty'}
-
-        self.command_history.append(command_str)
-
-        try:
-            result = self.parser.parse(command_str)
-            return result
-        except Exception as e:
-            return {'type': 'error', 'message': f'Parse error: {str(e)}'}
-
-
-class AdminCLI:
-    def __init__(self):
-        self.parser = AdminCommandParser()
-        self.is_interactive = False
-        self.admin_account = "admin@ragflow.io"
-        self.admin_password: str = "admin"
-        self.host: str = ""
-        self.port: int = 0
-
-    def verify_admin(self, args):
-
-        conn_info = self._parse_connection_args(args)
-        if 'error' in conn_info:
-            print(f"Error: {conn_info['error']}")
-            return
-
-        self.host = conn_info['host']
-        self.port = conn_info['port']
-        print(f"Attempt to access ip: {self.host}, port: {self.port}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/auth'
-
-        try_count = 0
-        while True:
-            try_count += 1
-            if try_count > 3:
-                return False
-
-            admin_passwd = input(f"password for {self.admin_account}: ").strip()
-            try:
-                self.admin_password = encode_to_base64(admin_passwd)
-                response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-                if response.status_code == 200:
-                    res_json = response.json()
-                    error_code = res_json.get('code', -1)
-                    if error_code == 0:
-                        print("Authentication successful.")
-                        return True
-                    else:
-                        error_message = res_json.get('message', 'Unknown error')
-                        print(f"Authentication failed: {error_message}, try again")
-                        continue
-                else:
-                    print(f"Bad response，status: {response.status_code}, try again")
-            except Exception:
-                print(f"Can't access {self.host}, port: {self.port}")
-
-    def _print_table_simple(self, data):
-        if not data:
-            print("No data to print")
-            return
-        if isinstance(data, dict):
-            # handle single row data
-            data = [data]
-
-        columns = list(data[0].keys())
-        col_widths = {}
-
-        for col in columns:
-            max_width = len(str(col))
-            for item in data:
-                value_len = len(str(item.get(col, '')))
-                if value_len > max_width:
-                    max_width = value_len
-            col_widths[col] = max(2, max_width)
-
-        # Generate delimiter
-        separator = "+" + "+".join(["-" * (col_widths[col] + 2) for col in columns]) + "+"
-
-        # Print header
-        print(separator)
-        header = "|" + "|".join([f" {col:<{col_widths[col]}} " for col in columns]) + "|"
-        print(header)
-        print(separator)
-
-        # Print data
-        for item in data:
-            row = "|"
-            for col in columns:
-                value = str(item.get(col, ''))
-                if len(value) > col_widths[col]:
-                    value = value[:col_widths[col] - 3] + "..."
-                row += f" {value:<{col_widths[col]}} |"
-            print(row)
-
-        print(separator)
-
-    def run_interactive(self):
-
-        self.is_interactive = True
-        print("RAGFlow Admin command line interface - Type '\\?' for help, '\\q' to quit")
-
-        while True:
-            try:
-                command = input("admin> ").strip()
-                if not command:
-                    continue
-
-                print(f"command: {command}")
-                result = self.parser.parse_command(command)
-                self.execute_command(result)
-
-                if isinstance(result, Tree):
-                    continue
-
-                if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
-                    break
-
-            except KeyboardInterrupt:
-                print("\nUse '\\q' to quit")
-            except EOFError:
-                print("\nGoodbye!")
-                break
-
-    def run_single_command(self, args):
-        conn_info = self._parse_connection_args(args)
-        if 'error' in conn_info:
-            print(f"Error: {conn_info['error']}")
-            return
-
-    def _parse_connection_args(self, args: List[str]) -> Dict[str, Any]:
-        parser = argparse.ArgumentParser(description='Admin CLI Client', add_help=False)
-        parser.add_argument('-h', '--host', default='localhost', help='Admin service host')
-        parser.add_argument('-p', '--port', type=int, default=8080, help='Admin service port')
-
-        try:
-            parsed_args, remaining_args = parser.parse_known_args(args)
-            return {
-                'host': parsed_args.host,
-                'port': parsed_args.port,
-            }
-        except SystemExit:
-            return {'error': 'Invalid connection arguments'}
-
-    def execute_command(self, parsed_command: Dict[str, Any]):
-
-        command_dict: dict
-        if isinstance(parsed_command, Tree):
-            command_dict = parsed_command.children[0]
-        else:
-            if parsed_command['type'] == 'error':
-                print(f"Error: {parsed_command['message']}")
-                return
-            else:
-                command_dict = parsed_command
-
-        # print(f"Parsed command: {command_dict}")
-
-        command_type = command_dict['type']
-
-        match command_type:
-            case 'list_services':
-                self._handle_list_services(command_dict)
-            case 'show_service':
-                self._handle_show_service(command_dict)
-            case 'restart_service':
-                self._handle_restart_service(command_dict)
-            case 'shutdown_service':
-                self._handle_shutdown_service(command_dict)
-            case 'startup_service':
-                self._handle_startup_service(command_dict)
-            case 'list_users':
-                self._handle_list_users(command_dict)
-            case 'show_user':
-                self._handle_show_user(command_dict)
-            case 'drop_user':
-                self._handle_drop_user(command_dict)
-            case 'alter_user':
-                self._handle_alter_user(command_dict)
-            case 'create_user':
-                self._handle_create_user(command_dict)
-            case 'activate_user':
-                self._handle_activate_user(command_dict)
-            case 'list_datasets':
-                self._handle_list_datasets(command_dict)
-            case 'list_agents':
-                self._handle_list_agents(command_dict)
-            case 'meta':
-                self._handle_meta_command(command_dict)
-            case _:
-                print(f"Command '{command_type}' would be executed with API")
-
-    def _handle_list_services(self, command):
-        print("Listing all services")
-
-        url = f'http://{self.host}:{self.port}/api/v1/admin/services'
-        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_show_service(self, command):
-        service_id: int = command['number']
-        print(f"Showing service: {service_id}")
-
-    def _handle_restart_service(self, command):
-        service_id: int = command['number']
-        print(f"Restart service {service_id}")
-
-    def _handle_shutdown_service(self, command):
-        service_id: int = command['number']
-        print(f"Shutdown service {service_id}")
-
-    def _handle_startup_service(self, command):
-        service_id: int = command['number']
-        print(f"Startup service {service_id}")
-
-    def _handle_list_users(self, command):
-        print("Listing all users")
-
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
-        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_show_user(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        print(f"Showing user: {username}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}'
-        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to get user {username}, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_drop_user(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        print(f"Drop user: {username}")
-
-    def _handle_alter_user(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        password_tree: Tree = command['password']
-        password: str = password_tree.children[0].strip("'\"")
-        print(f"Alter user: {username}, password: {password}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/password'
-        response = requests.put(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password), json={'new_password': encrypt(password)})
-        res_json = response.json()
-        if response.status_code == 200:
-            print(res_json["message"])
-        else:
-            print(f"Fail to alter password, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_create_user(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        password_tree: Tree = command['password']
-        password: str = password_tree.children[0].strip("'\"")
-        role: str = command['role']
-        print(f"Create user: {username}, password: {password}, role: {role}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
-        response = requests.post(
-            url,
-            auth=HTTPBasicAuth(self.admin_account, self.admin_password),
-            json={'username': username, 'password': encrypt(password), 'role': role}
-        )
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to create user {username}, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_activate_user(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        activate_tree: Tree = command['activate_status']
-        activate_status: str = activate_tree.children[0].strip("'\"")
-        if activate_status.lower() in ['on', 'off']:
-            print(f"Alter user {username} activate status, turn {activate_status.lower()}.")
-            url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/activate'
-            response = requests.put(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password), json={'activate_status': activate_status})
-            res_json = response.json()
-            if response.status_code == 200:
-                print(res_json["message"])
-            else:
-                print(f"Fail to alter activate status, code: {res_json['code']}, message: {res_json['message']}")
-        else:
-            print(f"Unknown activate status: {activate_status}.")
-
-    def _handle_list_datasets(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        print(f"Listing all datasets of user: {username}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/datasets'
-        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to get all datasets of {username}, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_list_agents(self, command):
-        username_tree: Tree = command['username']
-        username: str = username_tree.children[0].strip("'\"")
-        print(f"Listing all agents of user: {username}")
-        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{username}/agents'
-        response = requests.get(url, auth=HTTPBasicAuth(self.admin_account, self.admin_password))
-        res_json = response.json()
-        if response.status_code == 200:
-            self._print_table_simple(res_json['data'])
-        else:
-            print(f"Fail to get all agents of {username}, code: {res_json['code']}, message: {res_json['message']}")
-
-    def _handle_meta_command(self, command):
-        meta_command = command['command']
-        args = command.get('args', [])
-
-        if meta_command in ['?', 'h', 'help']:
-            self.show_help()
-        elif meta_command in ['q', 'quit', 'exit']:
-            print("Goodbye!")
-        else:
-            print(f"Meta command '{meta_command}' with args {args}")
-
-    def show_help(self):
-        """Help info"""
-        help_text = """
-Commands:
-  LIST SERVICES
-  SHOW SERVICE <service>
-  STARTUP SERVICE <service>
-  SHUTDOWN SERVICE <service>
-  RESTART SERVICE <service>
-  LIST USERS
-  SHOW USER <user>
-  DROP USER <user>
-  CREATE USER <user> <password>
-  ALTER USER PASSWORD <user> <new_password>
-  LIST DATASETS OF <user>
-  LIST AGENTS OF <user>
-
-Meta Commands:
-  \\?, \\h, \\help     Show this help
-  \\q, \\quit, \\exit   Quit the CLI
-        """
-        print(help_text)
-
-
-def main():
-    import sys
-
-    cli = AdminCLI()
-
-    if len(sys.argv) == 1 or (len(sys.argv) > 1 and sys.argv[1] == '-'):
-        print(r"""
-            ____  ___   ______________                 ___       __          _     
-           / __ \/   | / ____/ ____/ /___ _      __   /   | ____/ /___ ___  (_)___ 
-          / /_/ / /| |/ / __/ /_  / / __ \ | /| / /  / /| |/ __  / __ `__ \/ / __ \
-         / _, _/ ___ / /_/ / __/ / / /_/ / |/ |/ /  / ___ / /_/ / / / / / / / / / /
-        /_/ |_/_/  |_\____/_/   /_/\____/|__/|__/  /_/  |_\__,_/_/ /_/ /_/_/_/ /_/ 
-        """)
-        if cli.verify_admin(sys.argv):
-            cli.run_interactive()
-    else:
-        if cli.verify_admin(sys.argv):
-            cli.run_interactive()
-            # cli.run_single_command(sys.argv[1:])
-
-
-if __name__ == '__main__':
-    main()

admin/auth.py

@@ -1,57 +0,0 @@
-import logging
-import uuid
-from functools import wraps
-from flask import request, jsonify
-
-from exceptions import AdminException
-from api.db.init_data import encode_to_base64
-from api.db.services import UserService
-
-
-def check_admin(username: str, password: str):
-    users = UserService.query(email=username)
-    if not users:
-        logging.info(f"Username: {username} is not registered!")
-        user_info = {
-            "id": uuid.uuid1().hex,
-            "password": encode_to_base64("admin"),
-            "nickname": "admin",
-            "is_superuser": True,
-            "email": "admin@ragflow.io",
-            "creator": "system",
-            "status": "1",
-        }
-        if not UserService.save(**user_info):
-            raise AdminException("Can't init admin.", 500)
-
-    user = UserService.query_user(username, password)
-    if user:
-        return True
-    else:
-        return False
-
-
-def login_verify(f):
-    @wraps(f)
-    def decorated(*args, **kwargs):
-        auth = request.authorization
-        if not auth or 'username' not in auth.parameters or 'password' not in auth.parameters:
-            return jsonify({
-                "code": 401,
-                "message": "Authentication required",
-                "data": None
-            }), 200
-
-        username = auth.parameters['username']
-        password = auth.parameters['password']
-        # TODO: to check the username and password from DB
-        if check_admin(username, password) is False:
-            return jsonify({
-                "code": 403,
-                "message": "Access denied",
-                "data": None
-            }), 200
-
-        return f(*args, **kwargs)
-
-    return decorated

admin/build_cli_release.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+
+set -e
+
+echo "🚀 Start building..."
+echo "================================"
+
+PROJECT_NAME="ragflow-cli"
+
+RELEASE_DIR="release"
+BUILD_DIR="dist"
+SOURCE_DIR="src"
+PACKAGE_DIR="ragflow_cli"
+
+echo "🧹 Clean old build folder..."
+rm -rf release/
+
+echo "📁 Prepare source code..."
+mkdir release/$PROJECT_NAME/$SOURCE_DIR -p
+cp pyproject.toml release/$PROJECT_NAME/pyproject.toml
+cp README.md release/$PROJECT_NAME/README.md
+
+mkdir release/$PROJECT_NAME/$SOURCE_DIR/$PACKAGE_DIR -p
+cp admin_client.py release/$PROJECT_NAME/$SOURCE_DIR/$PACKAGE_DIR/admin_client.py
+
+if [ -d "release/$PROJECT_NAME/$SOURCE_DIR" ]; then
+    echo "✅ source dir: release/$PROJECT_NAME/$SOURCE_DIR"
+else
+    echo "❌ source dir not exist: release/$PROJECT_NAME/$SOURCE_DIR"
+    exit 1
+fi
+
+echo "🔨 Make build file..."
+cd release/$PROJECT_NAME
+export PYTHONPATH=$(pwd)
+python -m build
+
+echo "✅ check build result..."
+if [ -d "$BUILD_DIR" ]; then
+    echo "📦 Package generated:"
+    ls -la $BUILD_DIR/
+else
+    echo "❌ Build Failed: $BUILD_DIR not exist."
+    exit 1
+fi
+
+echo "🎉 Build finished successfully!"

admin/client/README.md

@@ -15,22 +15,55 @@ It consists of a server-side Service and a command-line client (CLI), both imple
 - **Admin Service**: A backend service that interfaces with the RAGFlow system to execute administrative operations and monitor its status.
 - **Admin CLI**: A command-line interface that allows users to connect to the Admin Service and issue commands for system management.
 
+
+
 ### Starting the Admin Service
 
-1.  Before start Admin Service, please make sure RAGFlow system is already started.
+#### Launching from source code
+
+1. Before start Admin Service, please make sure RAGFlow system is already started.
+
+2. Launch from source code:
+
+   ```bash
+   python admin/server/admin_server.py
+   ```
+   The service will start and listen for incoming connections from the CLI on the configured port. 
+
+#### Using docker image
+
+1. Before startup, please configure the `docker_compose.yml`  file to enable admin server:
+
+   ```bash
+   command:
+     - --enable-adminserver
+   ```
+
+2. Start the containers, the service will start and listen for incoming connections from the CLI on the configured port.
+
 
-2.  Run the service script:
-    ```bash
-    python admin/admin_server.py
-    ```
-    The service will start and listen for incoming connections from the CLI on the configured port.
 
 ### Using the Admin CLI
 
 1.  Ensure the Admin Service is running.
-2.  Launch the CLI client:
+2.  Install ragflow-cli.
     ```bash
-    python admin/admin_client.py -h 0.0.0.0 -p 9381
+    pip install ragflow-cli==0.21.1
+    ```
+3.  Launch the CLI client:
+    ```bash
+    ragflow-cli -h 127.0.0.1 -p 9381
+    ```
+    You will be prompted to enter the superuser's password to log in.
+    The default password is admin.
+
+    **Parameters:**
+    
+    - -h: RAGFlow admin server host address
+    
+    - -p: RAGFlow admin server port
+
+
 
 ## Supported Commands
 
@@ -42,12 +75,7 @@ Commands are case-insensitive and must be terminated with a semicolon (`;`).
     -   Lists all available services within the RAGFlow system.
 -   `SHOW SERVICE <id>;`
     -   Shows detailed status information for the service identified by `<id>`.
--   `STARTUP SERVICE <id>;`
+
-    -   Attempts to start the service identified by `<id>`.
--   `SHUTDOWN SERVICE <id>;`
-    -   Attempts to gracefully shut down the service identified by `<id>`.
--   `RESTART SERVICE <id>;`
-    -   Attempts to restart the service identified by `<id>`.
 
 ### User Management Commands
 
@@ -55,10 +83,17 @@ Commands are case-insensitive and must be terminated with a semicolon (`;`).
     -   Lists all users known to the system.
 -   `SHOW USER '<username>';`
     -   Shows details and permissions for the specified user. The username must be enclosed in single or double quotes.
+
+- `CREATE USER <username> <password>;`
+  - Create user by username and password. The username and password must be enclosed in single or double quotes.
+
 -   `DROP USER '<username>';`
     -   Removes the specified user from the system. Use with caution.
 -   `ALTER USER PASSWORD '<username>' '<new_password>';`
     -   Changes the password for the specified user.
+-   `ALTER USER ACTIVE <username> <on/off>;`
+    -   Changes the user to active or inactive.
+
 
 ### Data and Agent Commands
 

admin/client/admin_client.py

@@ -0,0 +1,931 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import argparse
+import base64
+from cmd import Cmd
+
+from Cryptodome.PublicKey import RSA
+from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
+from typing import Dict, List, Any
+from lark import Lark, Transformer, Tree
+import requests
+
+GRAMMAR = r"""
+start: command
+
+command: sql_command | meta_command
+
+sql_command: list_services
+           | show_service
+           | startup_service
+           | shutdown_service
+           | restart_service
+           | list_users
+           | show_user
+           | drop_user
+           | alter_user
+           | create_user
+           | activate_user
+           | list_datasets
+           | list_agents
+           | create_role
+           | drop_role
+           | alter_role
+           | list_roles
+           | show_role
+           | grant_permission
+           | revoke_permission
+           | alter_user_role
+           | show_user_permission
+
+// meta command definition
+meta_command: "\\" meta_command_name [meta_args]
+
+meta_command_name: /[a-zA-Z?]+/
+meta_args: (meta_arg)+
+
+meta_arg: /[^\\s"']+/ | quoted_string
+
+// command definition
+
+LIST: "LIST"i
+SERVICES: "SERVICES"i
+SHOW: "SHOW"i
+CREATE: "CREATE"i
+SERVICE: "SERVICE"i
+SHUTDOWN: "SHUTDOWN"i
+STARTUP: "STARTUP"i
+RESTART: "RESTART"i
+USERS: "USERS"i
+DROP: "DROP"i
+USER: "USER"i
+ALTER: "ALTER"i
+ACTIVE: "ACTIVE"i
+PASSWORD: "PASSWORD"i
+DATASETS: "DATASETS"i
+OF: "OF"i
+AGENTS: "AGENTS"i
+ROLE: "ROLE"i
+ROLES: "ROLES"i
+DESCRIPTION: "DESCRIPTION"i
+GRANT: "GRANT"i
+REVOKE: "REVOKE"i
+ALL: "ALL"i
+PERMISSION: "PERMISSION"i
+TO: "TO"i
+FROM: "FROM"i
+FOR: "FOR"i
+RESOURCES: "RESOURCES"i
+ON: "ON"i
+SET: "SET"i
+
+list_services: LIST SERVICES ";"
+show_service: SHOW SERVICE NUMBER ";"
+startup_service: STARTUP SERVICE NUMBER ";"
+shutdown_service: SHUTDOWN SERVICE NUMBER ";"
+restart_service: RESTART SERVICE NUMBER ";"
+
+list_users: LIST USERS ";"
+drop_user: DROP USER quoted_string ";"
+alter_user: ALTER USER PASSWORD quoted_string quoted_string ";"
+show_user: SHOW USER quoted_string ";"
+create_user: CREATE USER quoted_string quoted_string ";"
+activate_user: ALTER USER ACTIVE quoted_string status ";"
+
+list_datasets: LIST DATASETS OF quoted_string ";"
+list_agents: LIST AGENTS OF quoted_string ";"
+
+create_role: CREATE ROLE identifier [DESCRIPTION quoted_string] ";"
+drop_role: DROP ROLE identifier ";"
+alter_role: ALTER ROLE identifier SET DESCRIPTION quoted_string ";"
+list_roles: LIST ROLES ";"
+show_role: SHOW ROLE identifier ";"
+
+grant_permission: GRANT action_list ON identifier TO ROLE identifier ";"
+revoke_permission: REVOKE action_list ON identifier FROM ROLE identifier ";"
+alter_user_role: ALTER USER quoted_string SET ROLE identifier ";"
+show_user_permission: SHOW USER PERMISSION quoted_string ";"
+
+action_list: identifier ("," identifier)*
+
+identifier: WORD
+quoted_string: QUOTED_STRING
+status: WORD
+
+QUOTED_STRING: /'[^']+'/ | /"[^"]+"/
+WORD: /[a-zA-Z0-9_\-\.]+/
+NUMBER: /[0-9]+/
+
+%import common.WS
+%ignore WS
+"""
+
+
+class AdminTransformer(Transformer):
+
+    def start(self, items):
+        return items[0]
+
+    def command(self, items):
+        return items[0]
+
+    def list_services(self, items):
+        result = {'type': 'list_services'}
+        return result
+
+    def show_service(self, items):
+        service_id = int(items[2])
+        return {"type": "show_service", "number": service_id}
+
+    def startup_service(self, items):
+        service_id = int(items[2])
+        return {"type": "startup_service", "number": service_id}
+
+    def shutdown_service(self, items):
+        service_id = int(items[2])
+        return {"type": "shutdown_service", "number": service_id}
+
+    def restart_service(self, items):
+        service_id = int(items[2])
+        return {"type": "restart_service", "number": service_id}
+
+    def list_users(self, items):
+        return {"type": "list_users"}
+
+    def show_user(self, items):
+        user_name = items[2]
+        return {"type": "show_user", "user_name": user_name}
+
+    def drop_user(self, items):
+        user_name = items[2]
+        return {"type": "drop_user", "user_name": user_name}
+
+    def alter_user(self, items):
+        user_name = items[3]
+        new_password = items[4]
+        return {"type": "alter_user", "user_name": user_name, "password": new_password}
+
+    def create_user(self, items):
+        user_name = items[2]
+        password = items[3]
+        return {"type": "create_user", "user_name": user_name, "password": password, "role": "user"}
+
+    def activate_user(self, items):
+        user_name = items[3]
+        activate_status = items[4]
+        return {"type": "activate_user", "activate_status": activate_status, "user_name": user_name}
+
+    def list_datasets(self, items):
+        user_name = items[3]
+        return {"type": "list_datasets", "user_name": user_name}
+
+    def list_agents(self, items):
+        user_name = items[3]
+        return {"type": "list_agents", "user_name": user_name}
+
+    def create_role(self, items):
+        role_name = items[2]
+        if len(items) > 4:
+            description = items[4]
+            return {"type": "create_role", "role_name": role_name, "description": description}
+        else:
+            return {"type": "create_role", "role_name": role_name}
+
+    def drop_role(self, items):
+        role_name = items[2]
+        return {"type": "drop_role", "role_name": role_name}
+
+    def alter_role(self, items):
+        role_name = items[2]
+        description = items[5]
+        return {"type": "alter_role", "role_name": role_name, "description": description}
+
+    def list_roles(self, items):
+        return {"type": "list_roles"}
+
+    def show_role(self, items):
+        role_name = items[2]
+        return {"type": "show_role", "role_name": role_name}
+
+    def grant_permission(self, items):
+        action_list = items[1]
+        resource = items[3]
+        role_name = items[6]
+        return {"type": "grant_permission", "role_name": role_name, "resource": resource, "actions": action_list}
+
+    def revoke_permission(self, items):
+        action_list = items[1]
+        resource = items[3]
+        role_name = items[6]
+        return {
+            "type": "revoke_permission",
+            "role_name": role_name,
+            "resource": resource, "actions": action_list
+        }
+
+    def alter_user_role(self, items):
+        user_name = items[2]
+        role_name = items[5]
+        return {"type": "alter_user_role", "user_name": user_name, "role_name": role_name}
+
+    def show_user_permission(self, items):
+        user_name = items[3]
+        return {"type": "show_user_permission", "user_name": user_name}
+
+    def action_list(self, items):
+        return items
+
+    def meta_command(self, items):
+        command_name = str(items[0]).lower()
+        args = items[1:] if len(items) > 1 else []
+
+        # handle quoted parameter
+        parsed_args = []
+        for arg in args:
+            if hasattr(arg, 'value'):
+                parsed_args.append(arg.value)
+            else:
+                parsed_args.append(str(arg))
+
+        return {'type': 'meta', 'command': command_name, 'args': parsed_args}
+
+    def meta_command_name(self, items):
+        return items[0]
+
+    def meta_args(self, items):
+        return items
+
+
+def encrypt(input_string):
+    pub = '-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----'
+    pub_key = RSA.importKey(pub)
+    cipher = Cipher_pkcs1_v1_5.new(pub_key)
+    cipher_text = cipher.encrypt(base64.b64encode(input_string.encode('utf-8')))
+    return base64.b64encode(cipher_text).decode("utf-8")
+
+
+def encode_to_base64(input_string):
+    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
+    return base64_encoded.decode('utf-8')
+
+
+class AdminCLI(Cmd):
+    def __init__(self):
+        super().__init__()
+        self.parser = Lark(GRAMMAR, start='start', parser='lalr', transformer=AdminTransformer())
+        self.command_history = []
+        self.is_interactive = False
+        self.admin_account = "admin@ragflow.io"
+        self.admin_password: str = "admin"
+        self.session = requests.Session()
+        self.access_token: str = ""
+        self.host: str = ""
+        self.port: int = 0
+
+    intro = r"""Type "\h" for help."""
+    prompt = "admin> "
+
+    def onecmd(self, command: str) -> bool:
+        try:
+            result = self.parse_command(command)
+
+            if isinstance(result, dict):
+                if 'type' in result and result.get('type') == 'empty':
+                    return False
+
+            self.execute_command(result)
+
+            if isinstance(result, Tree):
+                return False
+
+            if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+                return True
+
+        except KeyboardInterrupt:
+            print("\nUse '\\q' to quit")
+        except EOFError:
+            print("\nGoodbye!")
+            return True
+        return False
+
+    def emptyline(self) -> bool:
+        return False
+
+    def default(self, line: str) -> bool:
+        return self.onecmd(line)
+
+    def parse_command(self, command_str: str) -> dict[str, str]:
+        if not command_str.strip():
+            return {'type': 'empty'}
+
+        self.command_history.append(command_str)
+
+        try:
+            result = self.parser.parse(command_str)
+            return result
+        except Exception as e:
+            return {'type': 'error', 'message': f'Parse error: {str(e)}'}
+
+    def verify_admin(self, arguments: dict, single_command: bool):
+        self.host = arguments['host']
+        self.port = arguments['port']
+        print(f"Attempt to access ip: {self.host}, port: {self.port}")
+        url = f"http://{self.host}:{self.port}/api/v1/admin/login"
+
+        attempt_count = 3
+        if single_command:
+            attempt_count = 1
+
+        try_count = 0
+        while True:
+            try_count += 1
+            if try_count > attempt_count:
+                return False
+
+            if single_command:
+                admin_passwd = arguments['password']
+            else:
+                admin_passwd = input(f"password for {self.admin_account}: ").strip()
+            try:
+                self.admin_password = encrypt(admin_passwd)
+                response = self.session.post(url, json={'email': self.admin_account, 'password': self.admin_password})
+                if response.status_code == 200:
+                    res_json = response.json()
+                    error_code = res_json.get('code', -1)
+                    if error_code == 0:
+                        self.session.headers.update({
+                            'Content-Type': 'application/json',
+                            'Authorization': response.headers['Authorization'],
+                            'User-Agent': 'RAGFlow-CLI/0.21.1'
+                        })
+                        print("Authentication successful.")
+                        return True
+                    else:
+                        error_message = res_json.get('message', 'Unknown error')
+                        print(f"Authentication failed: {error_message}, try again")
+                        continue
+                else:
+                    print(f"Bad response，status: {response.status_code}, password is wrong")
+            except Exception as e:
+                print(str(e))
+                print(f"Can't access {self.host}, port: {self.port}")
+
+    def _print_table_simple(self, data):
+        if not data:
+            print("No data to print")
+            return
+        if isinstance(data, dict):
+            # handle single row data
+            data = [data]
+
+        columns = list(data[0].keys())
+        col_widths = {}
+
+        def get_string_width(text):
+            half_width_chars = (
+                " !\"#$%&'()*+,-./0123456789:;<=>?@"
+                "ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`"
+                "abcdefghijklmnopqrstuvwxyz{|}~"
+                "\t\n\r"
+            )
+            width = 0
+            for char in text:
+                if char in half_width_chars:
+                    width += 1
+                else:
+                    width += 2
+            return width
+
+        for col in columns:
+            max_width = get_string_width(str(col))
+            for item in data:
+                value_len = get_string_width(str(item.get(col, '')))
+                if value_len > max_width:
+                    max_width = value_len
+            col_widths[col] = max(2, max_width)
+
+        # Generate delimiter
+        separator = "+" + "+".join(["-" * (col_widths[col] + 2) for col in columns]) + "+"
+
+        # Print header
+        print(separator)
+        header = "|" + "|".join([f" {col:<{col_widths[col]}} " for col in columns]) + "|"
+        print(header)
+        print(separator)
+
+        # Print data
+        for item in data:
+            row = "|"
+            for col in columns:
+                value = str(item.get(col, ''))
+                if get_string_width(value) > col_widths[col]:
+                    value = value[:col_widths[col] - 3] + "..."
+                row += f" {value:<{col_widths[col] - (get_string_width(value) - len(value))}} |"
+            print(row)
+
+        print(separator)
+
+    def run_interactive(self):
+
+        self.is_interactive = True
+        print("RAGFlow Admin command line interface - Type '\\?' for help, '\\q' to quit")
+
+        while True:
+            try:
+                command = input("admin> ").strip()
+                if not command:
+                    continue
+
+                print(f"command: {command}")
+                result = self.parse_command(command)
+                self.execute_command(result)
+
+                if isinstance(result, Tree):
+                    continue
+
+                if result.get('type') == 'meta' and result.get('command') in ['q', 'quit', 'exit']:
+                    break
+
+            except KeyboardInterrupt:
+                print("\nUse '\\q' to quit")
+            except EOFError:
+                print("\nGoodbye!")
+                break
+
+    def run_single_command(self, command: str):
+        result = self.parse_command(command)
+        self.execute_command(result)
+
+    def parse_connection_args(self, args: List[str]) -> Dict[str, Any]:
+        parser = argparse.ArgumentParser(description='Admin CLI Client', add_help=False)
+        parser.add_argument('-h', '--host', default='localhost', help='Admin service host')
+        parser.add_argument('-p', '--port', type=int, default=8080, help='Admin service port')
+        parser.add_argument('-w', '--password', default='admin', type=str, help='Superuser password')
+        parser.add_argument('command', nargs='?', help='Single command')
+        try:
+            parsed_args, remaining_args = parser.parse_known_args(args)
+            if remaining_args:
+                command = remaining_args[0]
+                return {
+                    'host': parsed_args.host,
+                    'port': parsed_args.port,
+                    'password': parsed_args.password,
+                    'command': command
+                }
+            else:
+                return {
+                    'host': parsed_args.host,
+                    'port': parsed_args.port,
+                }
+        except SystemExit:
+            return {'error': 'Invalid connection arguments'}
+
+    def execute_command(self, parsed_command: Dict[str, Any]):
+
+        command_dict: dict
+        if isinstance(parsed_command, Tree):
+            command_dict = parsed_command.children[0]
+        else:
+            if parsed_command['type'] == 'error':
+                print(f"Error: {parsed_command['message']}")
+                return
+            else:
+                command_dict = parsed_command
+
+        # print(f"Parsed command: {command_dict}")
+
+        command_type = command_dict['type']
+
+        match command_type:
+            case 'list_services':
+                self._handle_list_services(command_dict)
+            case 'show_service':
+                self._handle_show_service(command_dict)
+            case 'restart_service':
+                self._handle_restart_service(command_dict)
+            case 'shutdown_service':
+                self._handle_shutdown_service(command_dict)
+            case 'startup_service':
+                self._handle_startup_service(command_dict)
+            case 'list_users':
+                self._handle_list_users(command_dict)
+            case 'show_user':
+                self._handle_show_user(command_dict)
+            case 'drop_user':
+                self._handle_drop_user(command_dict)
+            case 'alter_user':
+                self._handle_alter_user(command_dict)
+            case 'create_user':
+                self._handle_create_user(command_dict)
+            case 'activate_user':
+                self._handle_activate_user(command_dict)
+            case 'list_datasets':
+                self._handle_list_datasets(command_dict)
+            case 'list_agents':
+                self._handle_list_agents(command_dict)
+            case 'create_role':
+                self._create_role(command_dict)
+            case 'drop_role':
+                self._drop_role(command_dict)
+            case 'alter_role':
+                self._alter_role(command_dict)
+            case 'list_roles':
+                self._list_roles(command_dict)
+            case 'show_role':
+                self._show_role(command_dict)
+            case 'grant_permission':
+                self._grant_permission(command_dict)
+            case 'revoke_permission':
+                self._revoke_permission(command_dict)
+            case 'alter_user_role':
+                self._alter_user_role(command_dict)
+            case 'show_user_permission':
+                self._show_user_permission(command_dict)
+            case 'meta':
+                self._handle_meta_command(command_dict)
+            case _:
+                print(f"Command '{command_type}' would be executed with API")
+
+    def _handle_list_services(self, command):
+        print("Listing all services")
+
+        url = f'http://{self.host}:{self.port}/api/v1/admin/services'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all services, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_show_service(self, command):
+        service_id: int = command['number']
+        print(f"Showing service: {service_id}")
+
+        url = f'http://{self.host}:{self.port}/api/v1/admin/services/{service_id}'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            res_data = res_json['data']
+            if 'status' in res_data and res_data['status'] == 'alive':
+                print(f"Service {res_data['service_name']} is alive, ")
+                if isinstance(res_data['message'], str):
+                    print(res_data['message'])
+                else:
+                    self._print_table_simple(res_data['message'])
+            else:
+                print(f"Service {res_data['service_name']} is down, {res_data['message']}")
+        else:
+            print(f"Fail to show service, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_restart_service(self, command):
+        service_id: int = command['number']
+        print(f"Restart service {service_id}")
+
+    def _handle_shutdown_service(self, command):
+        service_id: int = command['number']
+        print(f"Shutdown service {service_id}")
+
+    def _handle_startup_service(self, command):
+        service_id: int = command['number']
+        print(f"Startup service {service_id}")
+
+    def _handle_list_users(self, command):
+        print("Listing all users")
+
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all users, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_show_user(self, command):
+        username_tree: Tree = command['user_name']
+        user_name: str = username_tree.children[0].strip("'\"")
+        print(f"Showing user: {user_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_drop_user(self, command):
+        username_tree: Tree = command['user_name']
+        user_name: str = username_tree.children[0].strip("'\"")
+        print(f"Drop user: {user_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}'
+        response = self.session.delete(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            print(f"Fail to drop user, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_alter_user(self, command):
+        user_name_tree: Tree = command['user_name']
+        user_name: str = user_name_tree.children[0].strip("'\"")
+        password_tree: Tree = command['password']
+        password: str = password_tree.children[0].strip("'\"")
+        print(f"Alter user: {user_name}, password: {password}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/password'
+        response = self.session.put(url, json={'new_password': encrypt(password)})
+        res_json = response.json()
+        if response.status_code == 200:
+            print(res_json["message"])
+        else:
+            print(f"Fail to alter password, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_create_user(self, command):
+        user_name_tree: Tree = command['user_name']
+        user_name: str = user_name_tree.children[0].strip("'\"")
+        password_tree: Tree = command['password']
+        password: str = password_tree.children[0].strip("'\"")
+        role: str = command['role']
+        print(f"Create user: {user_name}, password: {password}, role: {role}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users'
+        response = self.session.post(
+            url,
+            json={'user_name': user_name, 'password': encrypt(password), 'role': role}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to create user {user_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_activate_user(self, command):
+        user_name_tree: Tree = command['user_name']
+        user_name: str = user_name_tree.children[0].strip("'\"")
+        activate_tree: Tree = command['activate_status']
+        activate_status: str = activate_tree.children[0].strip("'\"")
+        if activate_status.lower() in ['on', 'off']:
+            print(f"Alter user {user_name} activate status, turn {activate_status.lower()}.")
+            url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/activate'
+            response = self.session.put(url, json={'activate_status': activate_status})
+            res_json = response.json()
+            if response.status_code == 200:
+                print(res_json["message"])
+            else:
+                print(f"Fail to alter activate status, code: {res_json['code']}, message: {res_json['message']}")
+        else:
+            print(f"Unknown activate status: {activate_status}.")
+
+    def _handle_list_datasets(self, command):
+        username_tree: Tree = command['user_name']
+        user_name: str = username_tree.children[0].strip("'\"")
+        print(f"Listing all datasets of user: {user_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/datasets'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all datasets of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_list_agents(self, command):
+        username_tree: Tree = command['user_name']
+        user_name: str = username_tree.children[0].strip("'\"")
+        print(f"Listing all agents of user: {user_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name}/agents'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to get all agents of {user_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _create_role(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name: str = role_name_tree.children[0].strip("'\"")
+        desc_str: str = ''
+        if 'description' in command:
+            desc_tree: Tree = command['description']
+            desc_str = desc_tree.children[0].strip("'\"")
+
+        print(f"create role name: {role_name}, description: {desc_str}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles'
+        response = self.session.post(
+            url,
+            json={'role_name': role_name, 'description': desc_str}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to create role {role_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _drop_role(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name: str = role_name_tree.children[0].strip("'\"")
+        print(f"drop role name: {role_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}'
+        response = self.session.delete(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to drop role {role_name}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _alter_role(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name: str = role_name_tree.children[0].strip("'\"")
+        desc_tree: Tree = command['description']
+        desc_str: str = desc_tree.children[0].strip("'\"")
+
+        print(f"alter role name: {role_name}, description: {desc_str}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}'
+        response = self.session.put(
+            url,
+            json={'description': desc_str}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(
+                f"Fail to update role {role_name} with description: {desc_str}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _list_roles(self, command):
+        print("Listing all roles")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to list roles, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _show_role(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name: str = role_name_tree.children[0].strip("'\"")
+        print(f"show role: {role_name}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name}/permission'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(f"Fail to list roles, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _grant_permission(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name_str: str = role_name_tree.children[0].strip("'\"")
+        resource_tree: Tree = command['resource']
+        resource_str: str = resource_tree.children[0].strip("'\"")
+        action_tree_list: list = command['actions']
+        actions: list = []
+        for action_tree in action_tree_list:
+            action_str: str = action_tree.children[0].strip("'\"")
+            actions.append(action_str)
+        print(f"grant role_name: {role_name_str}, resource: {resource_str}, actions: {actions}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission'
+        response = self.session.post(
+            url,
+            json={'actions': actions, 'resource': resource_str}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(
+                f"Fail to grant role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _revoke_permission(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name_str: str = role_name_tree.children[0].strip("'\"")
+        resource_tree: Tree = command['resource']
+        resource_str: str = resource_tree.children[0].strip("'\"")
+        action_tree_list: list = command['actions']
+        actions: list = []
+        for action_tree in action_tree_list:
+            action_str: str = action_tree.children[0].strip("'\"")
+            actions.append(action_str)
+        print(f"revoke role_name: {role_name_str}, resource: {resource_str}, actions: {actions}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/roles/{role_name_str}/permission'
+        response = self.session.delete(
+            url,
+            json={'actions': actions, 'resource': resource_str}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(
+                f"Fail to revoke role {role_name_str} with {actions} on {resource_str}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _alter_user_role(self, command):
+        role_name_tree: Tree = command['role_name']
+        role_name_str: str = role_name_tree.children[0].strip("'\"")
+        user_name_tree: Tree = command['user_name']
+        user_name_str: str = user_name_tree.children[0].strip("'\"")
+        print(f"alter_user_role user_name: {user_name_str}, role_name: {role_name_str}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/role'
+        response = self.session.put(
+            url,
+            json={'role_name': role_name_str}
+        )
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(
+                f"Fail to alter user: {user_name_str} to role {role_name_str}, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _show_user_permission(self, command):
+        user_name_tree: Tree = command['user_name']
+        user_name_str: str = user_name_tree.children[0].strip("'\"")
+        print(f"show_user_permission user_name: {user_name_str}")
+        url = f'http://{self.host}:{self.port}/api/v1/admin/users/{user_name_str}/permission'
+        response = self.session.get(url)
+        res_json = response.json()
+        if response.status_code == 200:
+            self._print_table_simple(res_json['data'])
+        else:
+            print(
+                f"Fail to show user: {user_name_str} permission, code: {res_json['code']}, message: {res_json['message']}")
+
+    def _handle_meta_command(self, command):
+        meta_command = command['command']
+        args = command.get('args', [])
+
+        if meta_command in ['?', 'h', 'help']:
+            self.show_help()
+        elif meta_command in ['q', 'quit', 'exit']:
+            print("Goodbye!")
+        else:
+            print(f"Meta command '{meta_command}' with args {args}")
+
+    def show_help(self):
+        """Help info"""
+        help_text = """
+Commands:
+  LIST SERVICES
+  SHOW SERVICE <service>
+  STARTUP SERVICE <service>
+  SHUTDOWN SERVICE <service>
+  RESTART SERVICE <service>
+  LIST USERS
+  SHOW USER <user>
+  DROP USER <user>
+  CREATE USER <user> <password>
+  ALTER USER PASSWORD <user> <new_password>
+  ALTER USER ACTIVE <user> <on/off>
+  LIST DATASETS OF <user>
+  LIST AGENTS OF <user>
+
+Meta Commands:
+  \\?, \\h, \\help     Show this help
+  \\q, \\quit, \\exit   Quit the CLI
+        """
+        print(help_text)
+
+
+def main():
+    import sys
+
+    cli = AdminCLI()
+
+    args = cli.parse_connection_args(sys.argv)
+    if 'error' in args:
+        print(f"Error: {args['error']}")
+        return
+
+    if 'command' in args:
+        if 'password' not in args:
+            print("Error: password is missing")
+            return
+        if cli.verify_admin(args, single_command=True):
+            command: str = args['command']
+            print(f"Run single command: {command}")
+            cli.run_single_command(command)
+    else:
+        if cli.verify_admin(args, single_command=False):
+            print(r"""
+                ____  ___   ______________                 ___       __          _     
+               / __ \/   | / ____/ ____/ /___ _      __   /   | ____/ /___ ___  (_)___ 
+              / /_/ / /| |/ / __/ /_  / / __ \ | /| / /  / /| |/ __  / __ `__ \/ / __ \
+             / _, _/ ___ / /_/ / __/ / / /_/ / |/ |/ /  / ___ / /_/ / / / / / / / / / /
+            /_/ |_/_/  |_\____/_/   /_/\____/|__/|__/  /_/  |_\__,_/_/ /_/ /_/_/_/ /_/ 
+            """)
+            cli.cmdloop()
+
+
+if __name__ == '__main__':
+    main()

admin/client/pyproject.toml

@@ -0,0 +1,24 @@
+[project]
+name = "ragflow-cli"
+version = "0.21.1"
+description = "Admin Service's client of [RAGFlow](https://github.com/infiniflow/ragflow). The Admin Service provides user management and system monitoring. "
+authors = [{ name = "Lynn", email = "lynn_inf@hotmail.com" }]
+license = { text = "Apache License, Version 2.0" }
+readme = "README.md"
+requires-python = ">=3.10,<3.13"
+dependencies = [
+    "requests>=2.30.0,<3.0.0",
+    "beartype>=0.18.5,<0.19.0",
+    "pycryptodomex>=3.10.0",
+    "lark>=1.1.0",
+]
+
+[dependency-groups]
+test = [
+    "pytest>=8.3.5",
+    "requests>=2.32.3",
+    "requests-toolbelt>=1.0.0",
+]
+
+[project.scripts]
+ragflow-cli = "admin_client:main"

admin/responses.py

@@ -1,15 +0,0 @@
-from flask import jsonify
-
-def success_response(data=None, message="Success", code = 0):
-    return jsonify({
-        "code": code,
-        "message": message,
-        "data": data
-    }), 200
-
-def error_response(message="Error", code=-1, data=None):
-    return jsonify({
-        "code": code,
-        "message": message,
-        "data": data
-    }), 400

admin/routes.py

@@ -1,187 +0,0 @@
-from flask import Blueprint, request
-
-from auth import login_verify
-from responses import success_response, error_response
-from services import UserMgr, ServiceMgr, UserServiceMgr
-from exceptions import AdminException
-
-admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
-
-
-@admin_bp.route('/auth', methods=['GET'])
-@login_verify
-def auth_admin():
-    try:
-        return success_response(None, "Admin is authorized", 0)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/users', methods=['GET'])
-@login_verify
-def list_users():
-    try:
-        users = UserMgr.get_all_users()
-        return success_response(users, "Get all users", 0)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/users', methods=['POST'])
-@login_verify
-def create_user():
-    try:
-        data = request.get_json()
-        if not data or 'username' not in data or 'password' not in data:
-            return error_response("Username and password are required", 400)
-
-        username = data['username']
-        password = data['password']
-        role = data.get('role', 'user')
-
-        res = UserMgr.create_user(username, password, role)
-        if res["success"]:
-            user_info = res["user_info"]
-            user_info.pop("password") # do not return password
-            return success_response(user_info, "User created successfully")
-        else:
-            return error_response("create user failed")
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e))
-
-
-@admin_bp.route('/users/<username>', methods=['DELETE'])
-@login_verify
-def delete_user(username):
-    try:
-        UserMgr.delete_user(username)
-        return success_response(None, "User and all data deleted successfully")
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/users/<username>/password', methods=['PUT'])
-@login_verify
-def change_password(username):
-    try:
-        data = request.get_json()
-        if not data or 'new_password' not in data:
-            return error_response("New password is required", 400)
-
-        new_password = data['new_password']
-        msg = UserMgr.update_user_password(username, new_password)
-        return success_response(None, msg)
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/users/<username>/activate', methods=['PUT'])
-@login_verify
-def alter_user_activate_status(username):
-    try:
-        data = request.get_json()
-        if not data or 'activate_status' not in data:
-            return error_response("Activation status is required", 400)
-        activate_status = data['activate_status']
-        msg = UserMgr.update_user_activate_status(username, activate_status)
-        return success_response(None, msg)
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-@admin_bp.route('/users/<username>', methods=['GET'])
-@login_verify
-def get_user_details(username):
-    try:
-        user_details = UserMgr.get_user_details(username)
-        return success_response(user_details)
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-@admin_bp.route('/users/<username>/datasets', methods=['GET'])
-@login_verify
-def get_user_datasets(username):
-    try:
-        datasets_list = UserServiceMgr.get_user_datasets(username)
-        return success_response(datasets_list)
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/users/<username>/agents', methods=['GET'])
-@login_verify
-def get_user_agents(username):
-    try:
-        agents_list = UserServiceMgr.get_user_agents(username)
-        return success_response(agents_list)
-
-    except AdminException as e:
-        return error_response(e.message, e.code)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/services', methods=['GET'])
-@login_verify
-def get_services():
-    try:
-        services = ServiceMgr.get_all_services()
-        return success_response(services, "Get all services", 0)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/service_types/<service_type>', methods=['GET'])
-@login_verify
-def get_services_by_type(service_type_str):
-    try:
-        services = ServiceMgr.get_services_by_type(service_type_str)
-        return success_response(services)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/services/<service_id>', methods=['GET'])
-@login_verify
-def get_service(service_id):
-    try:
-        services = ServiceMgr.get_service_details(service_id)
-        return success_response(services)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/services/<service_id>', methods=['DELETE'])
-@login_verify
-def shutdown_service(service_id):
-    try:
-        services = ServiceMgr.shutdown_service(service_id)
-        return success_response(services)
-    except Exception as e:
-        return error_response(str(e), 500)
-
-
-@admin_bp.route('/services/<service_id>', methods=['PUT'])
-@login_verify
-def restart_service(service_id):
-    try:
-        services = ServiceMgr.restart_service(service_id)
-        return success_response(services)
-    except Exception as e:
-        return error_response(str(e), 500)

admin/server/admin_server.py

@@ -1,3 +1,18 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
 
 import os
 import signal
@@ -12,6 +27,9 @@ from api.utils.log_utils import init_root_logger
 from api.constants import SERVICE_CONF
 from api import settings
 from config import load_configurations, SERVICE_CONFIGS
+from auth import init_default_admin, setup_auth
+from flask_session import Session
+from flask_login import LoginManager
 
 stop_event = threading.Event()
 
@@ -27,7 +45,17 @@ if __name__ == '__main__':
 
     app = Flask(__name__)
     app.register_blueprint(admin_bp)
+    app.config["SESSION_PERMANENT"] = False
+    app.config["SESSION_TYPE"] = "filesystem"
+    app.config["MAX_CONTENT_LENGTH"] = int(
+        os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
+    )
+    Session(app)
+    login_manager = LoginManager()
+    login_manager.init_app(app)
     settings.init_settings()
+    setup_auth(login_manager)
+    init_default_admin()
     SERVICE_CONFIGS.configs = load_configurations(SERVICE_CONF)
 
     try:

admin/server/auth.py

@@ -0,0 +1,193 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
+import logging
+import uuid
+from functools import wraps
+from datetime import datetime
+from flask import request, jsonify
+from flask_login import current_user, login_user
+from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
+
+from api import settings
+from api.common.exceptions import AdminException, UserNotFoundError
+from api.db.init_data import encode_to_base64
+from api.db.services import UserService
+from api.db import ActiveEnum, StatusEnum
+from api.utils.crypt import decrypt
+from api.utils import (
+    current_timestamp,
+    datetime_format,
+    get_format_time,
+    get_uuid,
+)
+from api.utils.api_utils import (
+    construct_response,
+)
+
+
+def setup_auth(login_manager):
+    @login_manager.request_loader
+    def load_user(web_request):
+        jwt = Serializer(secret_key=settings.SECRET_KEY)
+        authorization = web_request.headers.get("Authorization")
+        if authorization:
+            try:
+                access_token = str(jwt.loads(authorization))
+
+                if not access_token or not access_token.strip():
+                    logging.warning("Authentication attempt with empty access token")
+                    return None
+
+                # Access tokens should be UUIDs (32 hex characters)
+                if len(access_token.strip()) < 32:
+                    logging.warning(f"Authentication attempt with invalid token format: {len(access_token)} chars")
+                    return None
+
+                user = UserService.query(
+                    access_token=access_token, status=StatusEnum.VALID.value
+                )
+                if user:
+                    if not user[0].access_token or not user[0].access_token.strip():
+                        logging.warning(f"User {user[0].email} has empty access_token in database")
+                        return None
+                    return user[0]
+                else:
+                    return None
+            except Exception as e:
+                logging.warning(f"load_user got exception {e}")
+                return None
+        else:
+            return None
+
+
+def init_default_admin():
+    # Verify that at least one active admin user exists. If not, create a default one.
+    users = UserService.query(is_superuser=True)
+    if not users:
+        default_admin = {
+            "id": uuid.uuid1().hex,
+            "password": encode_to_base64("admin"),
+            "nickname": "admin",
+            "is_superuser": True,
+            "email": "admin@ragflow.io",
+            "creator": "system",
+            "status": "1",
+        }
+        if not UserService.save(**default_admin):
+            raise AdminException("Can't init admin.", 500)
+    elif not any([u.is_active == ActiveEnum.ACTIVE.value for u in users]):
+        raise AdminException("No active admin. Please update 'is_active' in db manually.", 500)
+
+
+def check_admin_auth(func):
+    @wraps(func)
+    def wrapper(*args, **kwargs):
+        user = UserService.filter_by_id(current_user.id)
+        if not user:
+            raise UserNotFoundError(current_user.email)
+        if not user.is_superuser:
+            raise AdminException("Not admin", 403)
+        if user.is_active == ActiveEnum.INACTIVE.value:
+            raise AdminException(f"User {current_user.email} inactive", 403)
+
+        return func(*args, **kwargs)
+
+    return wrapper
+
+
+def login_admin(email: str, password: str):
+    """
+    :param email: admin email
+    :param password: string before decrypt
+    """
+    users = UserService.query(email=email)
+    if not users:
+        raise UserNotFoundError(email)
+    psw = decrypt(password)
+    user = UserService.query_user(email, psw)
+    if not user:
+        raise AdminException("Email and password do not match!")
+    if not user.is_superuser:
+        raise AdminException("Not admin", 403)
+    if user.is_active == ActiveEnum.INACTIVE.value:
+        raise AdminException(f"User {email} inactive", 403)
+
+    resp = user.to_json()
+    user.access_token = get_uuid()
+    login_user(user)
+    user.update_time = (current_timestamp(),)
+    user.update_date = (datetime_format(datetime.now()),)
+    user.last_login_time = get_format_time()
+    user.save()
+    msg = "Welcome back!"
+    return construct_response(data=resp, auth=user.get_id(), message=msg)
+
+
+def check_admin(username: str, password: str):
+    users = UserService.query(email=username)
+    if not users:
+        logging.info(f"Username: {username} is not registered!")
+        user_info = {
+            "id": uuid.uuid1().hex,
+            "password": encode_to_base64("admin"),
+            "nickname": "admin",
+            "is_superuser": True,
+            "email": "admin@ragflow.io",
+            "creator": "system",
+            "status": "1",
+        }
+        if not UserService.save(**user_info):
+            raise AdminException("Can't init admin.", 500)
+
+    user = UserService.query_user(username, password)
+    if user:
+        return True
+    else:
+        return False
+
+
+def login_verify(f):
+    @wraps(f)
+    def decorated(*args, **kwargs):
+        auth = request.authorization
+        if not auth or 'username' not in auth.parameters or 'password' not in auth.parameters:
+            return jsonify({
+                "code": 401,
+                "message": "Authentication required",
+                "data": None
+            }), 200
+
+        username = auth.parameters['username']
+        password = auth.parameters['password']
+        try:
+            if check_admin(username, password) is False:
+                return jsonify({
+                    "code": 500,
+                    "message": "Access denied",
+                    "data": None
+                }), 200
+        except Exception as e:
+            error_msg = str(e)
+            return jsonify({
+                "code": 500,
+                "message": error_msg
+            }), 200
+
+        return f(*args, **kwargs)
+
+    return decorated

admin/server/config.py

@@ -1,14 +1,33 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
 import logging
 import threading
 from enum import Enum
 
 from pydantic import BaseModel
 from typing import Any
-from api.utils import read_config
+from api.utils.configs import read_config
 from urllib.parse import urlparse
 
 
 class ServiceConfigs:
+    configs = dict
+
     def __init__(self):
         self.configs = []
         self.lock = threading.Lock()
@@ -32,9 +51,11 @@ class BaseConfig(BaseModel):
     host: str
     port: int
     service_type: str
+    detail_func_name: str
 
     def to_dict(self) -> dict[str, Any]:
-        return {'id': self.id, 'name': self.name, 'host': self.host, 'port': self.port, 'service_type': self.service_type}
+        return {'id': self.id, 'name': self.name, 'host': self.host, 'port': self.port,
+                'service_type': self.service_type}
 
 
 class MetaConfig(BaseConfig):
@@ -209,7 +230,9 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 name: str = f'ragflow_{ragflow_count}'
                 host: str = v['host']
                 http_port: int = v['http_port']
-                config = RAGFlowServerConfig(id=id_count, name=name, host=host, port=http_port, service_type="ragflow_server")
+                config = RAGFlowServerConfig(id=id_count, name=name, host=host, port=http_port,
+                                             service_type="ragflow_server",
+                                             detail_func_name="check_ragflow_server_alive")
                 configurations.append(config)
                 id_count += 1
             case "es":
@@ -222,7 +245,8 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 password: str = v.get('password')
                 config = ElasticsearchConfig(id=id_count, name=name, host=host, port=port, service_type="retrieval",
                                              retrieval_type="elasticsearch",
-                                             username=username, password=password)
+                                             username=username, password=password,
+                                             detail_func_name="get_es_cluster_stats")
                 configurations.append(config)
                 id_count += 1
 
@@ -233,8 +257,9 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 host = parts[0]
                 port = int(parts[1])
                 database: str = v.get('db_name', 'default_db')
-                config = InfinityConfig(id=id_count, name=name, host=host, port=port, service_type="retrieval", retrieval_type="infinity",
+                config = InfinityConfig(id=id_count, name=name, host=host, port=port, service_type="retrieval",
-                                        db_name=database)
+                                        retrieval_type="infinity",
+                                        db_name=database, detail_func_name="get_infinity_status")
                 configurations.append(config)
                 id_count += 1
             case "minio":
@@ -245,8 +270,9 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 port = int(parts[1])
                 user = v.get('user')
                 password = v.get('password')
-                config = MinioConfig(id=id_count, name=name, host=host, port=port, user=user, password=password, service_type="file_store",
+                config = MinioConfig(id=id_count, name=name, host=host, port=port, user=user, password=password,
-                                     store_type="minio")
+                                     service_type="file_store",
+                                     store_type="minio", detail_func_name="check_minio_alive")
                 configurations.append(config)
                 id_count += 1
             case "redis":
@@ -258,7 +284,7 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 password = v.get('password')
                 db: int = v.get('db')
                 config = RedisConfig(id=id_count, name=name, host=host, port=port, password=password, database=db,
-                                     service_type="message_queue", mq_type="redis")
+                                     service_type="message_queue", mq_type="redis", detail_func_name="get_redis_info")
                 configurations.append(config)
                 id_count += 1
             case "mysql":
@@ -268,7 +294,7 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                 username = v.get('user')
                 password = v.get('password')
                 config = MySQLConfig(id=id_count, name=name, host=host, port=port, username=username, password=password,
-                                     service_type="meta_data", meta_type="mysql")
+                                     service_type="meta_data", meta_type="mysql", detail_func_name="get_mysql_status")
                 configurations.append(config)
                 id_count += 1
             case "admin":

admin/server/models.py

@@ -0,0 +1,15 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#

admin/server/responses.py

@@ -0,0 +1,34 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
+from flask import jsonify
+
+
+def success_response(data=None, message="Success", code=0):
+    return jsonify({
+        "code": code,
+        "message": message,
+        "data": data
+    }), 200
+
+
+def error_response(message="Error", code=-1, data=None):
+    return jsonify({
+        "code": code,
+        "message": message,
+        "data": data
+    }), 400

admin/server/roles.py

@@ -0,0 +1,76 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import logging
+
+from typing import Dict, Any
+
+from api.common.exceptions import AdminException
+
+
+class RoleMgr:
+    @staticmethod
+    def create_role(role_name: str, description: str):
+        error_msg = f"not implement: create role: {role_name}, description: {description}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def update_role_description(role_name: str, description: str) -> Dict[str, Any]:
+        error_msg = f"not implement: update role: {role_name} with description: {description}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def delete_role(role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: drop role: {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def list_roles() -> Dict[str, Any]:
+        error_msg = "not implement: list roles"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def get_role_permission(role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: show role {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def grant_role_permission(role_name: str, actions: list, resource: str) -> Dict[str, Any]:
+        error_msg = f"not implement: grant role {role_name} actions: {actions} on {resource}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def revoke_role_permission(role_name: str, actions: list, resource: str) -> Dict[str, Any]:
+        error_msg = f"not implement: revoke role {role_name} actions: {actions} on {resource}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def update_user_role(user_name: str, role_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: update user role: {user_name} to role {role_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)
+
+    @staticmethod
+    def get_user_permission(user_name: str) -> Dict[str, Any]:
+        error_msg = f"not implement: get user permission: {user_name}"
+        logging.error(error_msg)
+        raise AdminException(error_msg)

admin/server/routes.py

@@ -0,0 +1,371 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import secrets
+
+from flask import Blueprint, request
+from flask_login import current_user, logout_user, login_required
+
+from auth import login_verify, login_admin, check_admin_auth
+from responses import success_response, error_response
+from services import UserMgr, ServiceMgr, UserServiceMgr
+from roles import RoleMgr
+from api.common.exceptions import AdminException
+
+admin_bp = Blueprint('admin', __name__, url_prefix='/api/v1/admin')
+
+
+@admin_bp.route('/login', methods=['POST'])
+def login():
+    if not request.json:
+        return error_response('Authorize admin failed.' ,400)
+    try:
+        email = request.json.get("email", "")
+        password = request.json.get("password", "")
+        return login_admin(email, password)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/logout', methods=['GET'])
+@login_required
+def logout():
+    try:
+        current_user.access_token = f"INVALID_{secrets.token_hex(16)}"
+        current_user.save()
+        logout_user()
+        return success_response(True)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/auth', methods=['GET'])
+@login_verify
+def auth_admin():
+    try:
+        return success_response(None, "Admin is authorized", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users', methods=['GET'])
+@login_required
+@check_admin_auth
+def list_users():
+    try:
+        users = UserMgr.get_all_users()
+        return success_response(users, "Get all users", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users', methods=['POST'])
+@login_required
+@check_admin_auth
+def create_user():
+    try:
+        data = request.get_json()
+        if not data or 'username' not in data or 'password' not in data:
+            return error_response("Username and password are required", 400)
+
+        username = data['username']
+        password = data['password']
+        role = data.get('role', 'user')
+
+        res = UserMgr.create_user(username, password, role)
+        if res["success"]:
+            user_info = res["user_info"]
+            user_info.pop("password")  # do not return password
+            return success_response(user_info, "User created successfully")
+        else:
+            return error_response("create user failed")
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e))
+
+
+@admin_bp.route('/users/<username>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def delete_user(username):
+    try:
+        res = UserMgr.delete_user(username)
+        if res["success"]:
+            return success_response(None, res["message"])
+        else:
+            return error_response(res["message"])
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/password', methods=['PUT'])
+@login_required
+@check_admin_auth
+def change_password(username):
+    try:
+        data = request.get_json()
+        if not data or 'new_password' not in data:
+            return error_response("New password is required", 400)
+
+        new_password = data['new_password']
+        msg = UserMgr.update_user_password(username, new_password)
+        return success_response(None, msg)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/activate', methods=['PUT'])
+@login_required
+@check_admin_auth
+def alter_user_activate_status(username):
+    try:
+        data = request.get_json()
+        if not data or 'activate_status' not in data:
+            return error_response("Activation status is required", 400)
+        activate_status = data['activate_status']
+        msg = UserMgr.update_user_activate_status(username, activate_status)
+        return success_response(None, msg)
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_details(username):
+    try:
+        user_details = UserMgr.get_user_details(username)
+        return success_response(user_details)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/datasets', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_datasets(username):
+    try:
+        datasets_list = UserServiceMgr.get_user_datasets(username)
+        return success_response(datasets_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<username>/agents', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_agents(username):
+    try:
+        agents_list = UserServiceMgr.get_user_agents(username)
+        return success_response(agents_list)
+
+    except AdminException as e:
+        return error_response(e.message, e.code)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_services():
+    try:
+        services = ServiceMgr.get_all_services()
+        return success_response(services, "Get all services", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/service_types/<service_type>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_services_by_type(service_type_str):
+    try:
+        services = ServiceMgr.get_services_by_type(service_type_str)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_service(service_id):
+    try:
+        services = ServiceMgr.get_service_details(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def shutdown_service(service_id):
+    try:
+        services = ServiceMgr.shutdown_service(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/services/<service_id>', methods=['PUT'])
+@login_required
+@check_admin_auth
+def restart_service(service_id):
+    try:
+        services = ServiceMgr.restart_service(service_id)
+        return success_response(services)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles', methods=['POST'])
+@login_required
+@check_admin_auth
+def create_role():
+    try:
+        data = request.get_json()
+        if not data or 'role_name' not in data:
+            return error_response("Role name is required", 400)
+        role_name: str = data['role_name']
+        description: str = data['description']
+        res = RoleMgr.create_role(role_name, description)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>', methods=['PUT'])
+@login_required
+@check_admin_auth
+def update_role(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'description' not in data:
+            return error_response("Role description is required", 400)
+        description: str = data['description']
+        res = RoleMgr.update_role_description(role_name, description)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def delete_role(role_name: str):
+    try:
+        res = RoleMgr.delete_role(role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles', methods=['GET'])
+@login_required
+@check_admin_auth
+def list_roles():
+    try:
+        res = RoleMgr.list_roles()
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_role_permission(role_name: str):
+    try:
+        res = RoleMgr.get_role_permission(role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['POST'])
+@login_required
+@check_admin_auth
+def grant_role_permission(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'actions' not in data or 'resource' not in data:
+            return error_response("Permission is required", 400)
+        actions: list = data['actions']
+        resource: str = data['resource']
+        res = RoleMgr.grant_role_permission(role_name, actions, resource)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/roles/<role_name>/permission', methods=['DELETE'])
+@login_required
+@check_admin_auth
+def revoke_role_permission(role_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'actions' not in data or 'resource' not in data:
+            return error_response("Permission is required", 400)
+        actions: list = data['actions']
+        resource: str = data['resource']
+        res = RoleMgr.revoke_role_permission(role_name, actions, resource)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<user_name>/role', methods=['PUT'])
+@login_required
+@check_admin_auth
+def update_user_role(user_name: str):
+    try:
+        data = request.get_json()
+        if not data or 'role_name' not in data:
+            return error_response("Role name is required", 400)
+        role_name: str = data['role_name']
+        res = RoleMgr.update_user_role(user_name, role_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route('/users/<user_name>/permission', methods=['GET'])
+@login_required
+@check_admin_auth
+def get_user_permission(user_name: str):
+    try:
+        res = RoleMgr.get_user_permission(user_name)
+        return success_response(res)
+    except Exception as e:
+        return error_response(str(e), 500)

admin/server/services.py

@@ -1,22 +1,48 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
 import re
 from werkzeug.security import check_password_hash
 from api.db import ActiveEnum
 from api.db.services import UserService
-from api.db.joint_services.user_account_service import create_new_user
+from api.db.joint_services.user_account_service import create_new_user, delete_user_data
 from api.db.services.canvas_service import UserCanvasService
 from api.db.services.user_service import TenantService
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.utils.crypt import decrypt
-from exceptions import AdminException, UserAlreadyExistsError, UserNotFoundError
+from api.utils import health_utils
+
+from api.common.exceptions import AdminException, UserAlreadyExistsError, UserNotFoundError
 from config import SERVICE_CONFIGS
 
+
 class UserMgr:
     @staticmethod
     def get_all_users():
         users = UserService.get_all_users()
         result = []
         for user in users:
-            result.append({'email': user.email, 'nickname': user.nickname, 'create_date': user.create_date, 'is_active': user.is_active})
+            result.append({
+                'email': user.email,
+                'nickname': user.nickname,
+                'create_date': user.create_date,
+                'is_active': user.is_active,
+                'is_superuser': user.is_superuser,
+            })
         return result
 
     @staticmethod
@@ -29,7 +55,6 @@ class UserMgr:
                 'email': user.email,
                 'language': user.language,
                 'last_login_time': user.last_login_time,
-                'is_authenticated': user.is_authenticated,
                 'is_active': user.is_active,
                 'is_anonymous': user.is_anonymous,
                 'login_channel': user.login_channel,
@@ -61,7 +86,13 @@ class UserMgr:
     @staticmethod
     def delete_user(username):
         # use email to delete
-        raise AdminException("delete_user: not implemented")
+        user_list = UserService.query_user_by_email(username)
+        if not user_list:
+            raise UserNotFoundError(username)
+        if len(user_list) > 1:
+            raise AdminException(f"Exist more than 1 user: {username}!")
+        usr = user_list[0]
+        return delete_user_data(usr.id)
 
     @staticmethod
     def update_user_password(username, new_password) -> str:
@@ -104,6 +135,7 @@ class UserMgr:
         UserService.update_user(usr.id, {"is_active": target_status})
         return f"Turn {_activate_status} user activate status successfully!"
 
+
 class UserServiceMgr:
 
     @staticmethod
@@ -134,7 +166,13 @@ class UserServiceMgr:
         tenants = TenantService.get_joined_tenants_by_user_id(usr.id)
         tenant_ids = [m["tenant_id"] for m in tenants]
         # filter permitted agents and owned agents
-        return UserCanvasService.get_all_agents_by_tenant_ids(tenant_ids, usr.id)
+        res = UserCanvasService.get_all_agents_by_tenant_ids(tenant_ids, usr.id)
+        return [{
+            'title': r['title'],
+            'permission': r['permission'],
+            'canvas_category': r['canvas_category'].split('_')[0]
+        } for r in res]
+
 
 class ServiceMgr:
 
@@ -142,8 +180,17 @@ class ServiceMgr:
     def get_all_services():
         result = []
         configs = SERVICE_CONFIGS.configs
-        for config in configs:
+        for service_id, config in enumerate(configs):
-            result.append(config.to_dict())
+            config_dict = config.to_dict()
+            try:
+                service_detail = ServiceMgr.get_service_details(service_id)
+                if "status" in service_detail:
+                    config_dict['status'] = service_detail['status']
+                else:
+                    config_dict['status'] = 'timeout'
+            except Exception:
+                config_dict['status'] = 'timeout'
+            result.append(config_dict)
         return result
 
     @staticmethod
@@ -152,7 +199,22 @@ class ServiceMgr:
 
     @staticmethod
     def get_service_details(service_id: int):
-        raise AdminException("get_service_details: not implemented")
+        service_id = int(service_id)
+        configs = SERVICE_CONFIGS.configs
+        service_config_mapping = {
+            c.id: {
+                'name': c.name,
+                'detail_func_name': c.detail_func_name
+            } for c in configs
+        }
+        service_info = service_config_mapping.get(service_id, {})
+        if not service_info:
+            raise AdminException(f"invalid service_id: {service_id}")
+
+        detail_func = getattr(health_utils, service_info.get('detail_func_name'))
+        res = detail_func()
+        res.update({'service_name': service_info.get('name')})
+        return res
 
     @staticmethod
     def shutdown_service(service_id: int):

agent/canvas.py

@@ -153,6 +153,16 @@ class Graph:
     def get_tenant_id(self):
         return self._tenant_id
 
+    def get_variable_value(self, exp: str) -> Any:
+        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
+        if exp.find("@") < 0:
+            return self.globals[exp]
+        cpn_id, var_nm = exp.split("@")
+        cpn = self.get_component(cpn_id)
+        if not cpn:
+            raise Exception(f"Can't find variable: '{cpn_id}@{var_nm}'")
+        return cpn["obj"].output(var_nm)
+
 
 class Canvas(Graph):
 
@@ -193,7 +203,6 @@ class Canvas(Graph):
             self.history = []
             self.retrieval = []
             self.memory = []
-
         for k in self.globals.keys():
             if isinstance(self.globals[k], str):
                 self.globals[k] = ""
@@ -282,7 +291,6 @@ class Canvas(Graph):
                     "thoughts": self.get_component_thoughts(self.path[i])
                 })
             _run_batch(idx, to)
-
             # post processing of components invocation
             for i in range(idx, to):
                 cpn = self.get_component(self.path[i])
@@ -383,7 +391,6 @@ class Canvas(Graph):
                 self.path = path
                 yield decorate("user_inputs", {"inputs": another_inputs, "tips": tips})
                 return
-
         self.path = self.path[:idx]
         if not self.error:
             yield decorate("workflow_finished",
@@ -406,16 +413,6 @@ class Canvas(Graph):
             return False
         return True
 
-    def get_variable_value(self, exp: str) -> Any:
-        exp = exp.strip("{").strip("}").strip(" ").strip("{").strip("}")
-        if exp.find("@") < 0:
-            return self.globals[exp]
-        cpn_id, var_nm = exp.split("@")
-        cpn = self.get_component(cpn_id)
-        if not cpn:
-            raise Exception(f"Can't find variable: '{cpn_id}@{var_nm}'")
-        return cpn["obj"].output(var_nm)
-
     def get_history(self, window_size):
         convs = []
         if window_size <= 0:

agent/component/agent_with_tools.py

@@ -137,7 +137,7 @@ class Agent(LLM, ToolBase):
             res.update(cpn.get_input_form())
         return res
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
     def _invoke(self, **kwargs):
         if kwargs.get("user_prompt"):
             usr_pmt = ""
@@ -346,3 +346,11 @@ Respond immediately with your final comprehensive answer.
 
         return "Error occurred."
 
+    def reset(self, temp=False):
+        """
+        Reset all tools if they have a reset method. This avoids errors for tools like MCPToolCallSession.
+        """
+        for k, cpn in self.tools.items():
+            if hasattr(cpn, "reset") and callable(cpn.reset):
+                cpn.reset()
+

agent/component/base.py

@@ -431,7 +431,7 @@ class ComponentBase(ABC):
         self.set_output("_elapsed_time", time.perf_counter() - self.output("_created_time"))
         return self.output()
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         raise NotImplementedError()
 

agent/component/invoke.py

@@ -19,11 +19,12 @@ import os
 import re
 import time
 from abc import ABC
+
 import requests
 
+from agent.component.base import ComponentBase, ComponentParamBase
 from api.utils.api_utils import timeout
 from deepdoc.parser import HtmlParser
-from agent.component.base import ComponentBase, ComponentParamBase
 
 
 class InvokeParam(ComponentParamBase):
@@ -43,17 +44,17 @@ class InvokeParam(ComponentParamBase):
         self.datatype = "json"  # New parameter to determine data posting type
 
     def check(self):
-        self.check_valid_value(self.method.lower(), "Type of content from the crawler", ['get', 'post', 'put'])
+        self.check_valid_value(self.method.lower(), "Type of content from the crawler", ["get", "post", "put"])
         self.check_empty(self.url, "End point URL")
         self.check_positive_integer(self.timeout, "Timeout time in second")
         self.check_boolean(self.clean_html, "Clean HTML")
-        self.check_valid_value(self.datatype.lower(), "Data post type", ['json', 'formdata'])  # Check for valid datapost value
+        self.check_valid_value(self.datatype.lower(), "Data post type", ["json", "formdata"])  # Check for valid datapost value
 
 
 class Invoke(ComponentBase, ABC):
     component_name = "Invoke"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         args = {}
         for para in self._param.variables:
@@ -63,6 +64,18 @@ class Invoke(ComponentBase, ABC):
                 args[para["key"]] = self._canvas.get_variable_value(para["ref"])
 
         url = self._param.url.strip()
+
+        def replace_variable(match):
+            var_name = match.group(1)
+            try:
+                value = self._canvas.get_variable_value(var_name)
+                return str(value or "")
+            except Exception:
+                return ""
+
+        # {base_url} or {component_id@variable_name}
+        url = re.sub(r"\{([a-zA-Z_][a-zA-Z0-9_.@-]*)\}", replace_variable, url)
+
         if url.find("http") != 0:
             url = "http://" + url
 
@@ -75,52 +88,32 @@ class Invoke(ComponentBase, ABC):
             proxies = {"http": self._param.proxy, "https": self._param.proxy}
 
         last_e = ""
-        for _ in range(self._param.max_retries+1):
+        for _ in range(self._param.max_retries + 1):
             try:
-                if method == 'get':
+                if method == "get":
-                    response = requests.get(url=url,
+                    response = requests.get(url=url, params=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                            params=args,
-                                            headers=headers,
-                                            proxies=proxies,
-                                            timeout=self._param.timeout)
                     if self._param.clean_html:
                         sections = HtmlParser()(None, response.content)
                         self.set_output("result", "\n".join(sections))
                     else:
                         self.set_output("result", response.text)
 
-                if method == 'put':
+                if method == "put":
-                    if self._param.datatype.lower() == 'json':
+                    if self._param.datatype.lower() == "json":
-                        response = requests.put(url=url,
+                        response = requests.put(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                json=args,
-                                                headers=headers,
-                                                proxies=proxies,
-                                                timeout=self._param.timeout)
                     else:
-                        response = requests.put(url=url,
+                        response = requests.put(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                data=args,
-                                                headers=headers,
-                                                proxies=proxies,
-                                                timeout=self._param.timeout)
                     if self._param.clean_html:
                         sections = HtmlParser()(None, response.content)
                         self.set_output("result", "\n".join(sections))
                     else:
                         self.set_output("result", response.text)
 
-                if method == 'post':
+                if method == "post":
-                    if self._param.datatype.lower() == 'json':
+                    if self._param.datatype.lower() == "json":
-                        response = requests.post(url=url,
+                        response = requests.post(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                 json=args,
-                                                 headers=headers,
-                                                 proxies=proxies,
-                                                 timeout=self._param.timeout)
                     else:
-                        response = requests.post(url=url,
+                        response = requests.post(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                                                 data=args,
-                                                 headers=headers,
-                                                 proxies=proxies,
-                                                 timeout=self._param.timeout)
                     if self._param.clean_html:
                         self.set_output("result", "\n".join(sections))
                     else:

agent/component/llm.py

@@ -101,6 +101,8 @@ class LLM(ComponentBase):
 
     def get_input_elements(self) -> dict[str, Any]:
         res = self.get_input_elements_from_text(self._param.sys_prompt)
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
         for prompt in self._param.prompts:
             d = self.get_input_elements_from_text(prompt["content"])
             res.update(d)
@@ -112,6 +114,17 @@ class LLM(ComponentBase):
     def add2system_prompt(self, txt):
         self._param.sys_prompt += txt
 
+    def _sys_prompt_and_msg(self, msg, args):
+        if isinstance(self._param.prompts, str):
+            self._param.prompts = [{"role": "user", "content": self._param.prompts}]
+        for p in self._param.prompts:
+            if msg and msg[-1]["role"] == p["role"]:
+                continue
+            p = deepcopy(p)
+            p["content"] = self.string_format(p["content"], args)
+            msg.append(p)
+        return msg, self.string_format(self._param.sys_prompt, args)
+
     def _prepare_prompt_variables(self):
         if self._param.visual_files_var:
             self.imgs = self._canvas.get_variable_value(self._param.visual_files_var)
@@ -127,7 +140,6 @@ class LLM(ComponentBase):
 
         args = {}
         vars = self.get_input_elements() if not self._param.debug_inputs else self._param.debug_inputs
-        sys_prompt = self._param.sys_prompt
         for k, o in vars.items():
             args[k] = o["value"]
             if not isinstance(args[k], str):
@@ -137,16 +149,8 @@ class LLM(ComponentBase):
                     args[k] = str(args[k])
             self.set_input_value(k, args[k])
 
-        msg = self._canvas.get_history(self._param.message_history_window_size)[:-1]
+        msg, sys_prompt = self._sys_prompt_and_msg(self._canvas.get_history(self._param.message_history_window_size)[:-1], args)
-        for p in self._param.prompts:
-            if msg and msg[-1]["role"] == p["role"]:
-                continue
-            msg.append(deepcopy(p))
-
-        sys_prompt = self.string_format(sys_prompt, args)
         user_defined_prompt, sys_prompt = self._extract_prompts(sys_prompt)
-        for m in msg:
-            m["content"] = self.string_format(m["content"], args)
         if self._param.cite and self._canvas.get_reference()["chunks"]:
             sys_prompt += citation_prompt(user_defined_prompt)
 
@@ -201,7 +205,7 @@ class LLM(ComponentBase):
             for txt in self.chat_mdl.chat_streamly(msg[0]["content"], msg[1:], self._param.gen_conf(), images=self.imgs, **kwargs):
                 yield delta(txt)
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         def clean_formated_answer(ans: str) -> str:
             ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)

agent/component/message.py

@@ -127,7 +127,7 @@ class Message(ComponentBase):
         ]
         return any([re.search(p, content) for p in patt])
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
     def _invoke(self, **kwargs):
         rand_cnt = random.choice(self._param.content)
         if self._param.stream and not self._is_jinjia2(rand_cnt):

agent/component/switch.py

@@ -61,7 +61,7 @@ class SwitchParam(ComponentParamBase):
 class Switch(ComponentBase, ABC):
     component_name = "Switch"
 
-    @timeout(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         for cond in self._param.conditions:
             res = []

agent/tools/code_exec.py

@@ -156,7 +156,7 @@ class CodeExec(ToolBase, ABC):
             self.set_output("_ERROR", "construct code request error: " + str(e))
 
         try:
-            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60))
+            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10*60)))
             logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
             if resp.status_code != 200:
                 resp.raise_for_status()

agent/tools/exesql.py

@@ -53,12 +53,13 @@ class ExeSQLParam(ToolParamBase):
         self.max_records = 1024
 
     def check(self):
-        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgres', 'mariadb', 'mssql'])
+        self.check_valid_value(self.db_type, "Choose DB type", ['mysql', 'postgres', 'mariadb', 'mssql', 'IBM DB2', 'trino'])
         self.check_empty(self.database, "Database name")
         self.check_empty(self.username, "database username")
         self.check_empty(self.host, "IP Address")
         self.check_positive_integer(self.port, "IP Port")
-        self.check_empty(self.password, "Database password")
+        if self.db_type != "trino":
+            self.check_empty(self.password, "Database password")
         self.check_positive_integer(self.max_records, "Maximum number of records")
         if self.database == "rag_flow":
             if self.host == "ragflow-mysql":
@@ -123,6 +124,94 @@ class ExeSQL(ToolBase, ABC):
                     r'PWD=' + self._param.password
             )
             db = pyodbc.connect(conn_str)
+        elif self._param.db_type == 'trino':
+            try:
+                import trino
+                from trino.auth import BasicAuthentication
+            except Exception:
+                raise Exception("Missing dependency 'trino'. Please install: pip install trino")
+
+            def _parse_catalog_schema(db: str):
+                if not db:
+                    return None, None
+                if "." in db:
+                    c, s = db.split(".", 1)
+                elif "/" in db:
+                    c, s = db.split("/", 1)
+                else:
+                    c, s = db, "default"
+                return c, s
+
+            catalog, schema = _parse_catalog_schema(self._param.database)
+            if not catalog:
+                raise Exception("For Trino, `database` must be 'catalog.schema' or at least 'catalog'.")
+
+            http_scheme = "https" if os.environ.get("TRINO_USE_TLS", "0") == "1" else "http"
+            auth = None
+            if http_scheme == "https" and self._param.password:
+                auth = BasicAuthentication(self._param.username, self._param.password)
+
+            try:
+                db = trino.dbapi.connect(
+                    host=self._param.host,
+                    port=int(self._param.port or 8080),
+                    user=self._param.username or "ragflow",
+                    catalog=catalog,
+                    schema=schema or "default",
+                    http_scheme=http_scheme,
+                    auth=auth
+                )
+            except Exception as e:
+                raise Exception("Database Connection Failed! \n" + str(e))
+        elif self._param.db_type == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={self._param.database};"
+                f"HOSTNAME={self._param.host};"
+                f"PORT={self._param.port};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={self._param.username};"
+                f"PWD={self._param.password};"
+            )
+            try:
+                conn = ibm_db.connect(conn_str, "", "")
+            except Exception as e:
+                raise Exception("Database Connection Failed! \n" + str(e))
+
+            sql_res = []
+            formalized_content = []
+            for single_sql in sqls:
+                single_sql = single_sql.replace("```", "").strip()
+                if not single_sql:
+                    continue
+                single_sql = re.sub(r"\[ID:[0-9]+\]", "", single_sql)
+
+                stmt = ibm_db.exec_immediate(conn, single_sql)
+                rows = []
+                row = ibm_db.fetch_assoc(stmt)
+                while row and len(rows) < self._param.max_records:
+                    rows.append(row)
+                    row = ibm_db.fetch_assoc(stmt)
+
+                if not rows:
+                    sql_res.append({"content": "No record in the database!"})
+                    continue
+
+                df = pd.DataFrame(rows)
+                for col in df.columns:
+                    if pd.api.types.is_datetime64_any_dtype(df[col]):
+                        df[col] = df[col].dt.strftime("%Y-%m-%d")
+
+                df = df.where(pd.notnull(df), None)
+
+                sql_res.append(convert_decimals(df.to_dict(orient="records")))
+                formalized_content.append(df.to_markdown(index=False, floatfmt=".6f"))
+
+            ibm_db.close(conn)
+
+            self.set_output("json", sql_res)
+            self.set_output("formalized_content", "\n\n".join(formalized_content))
+            return self.output("formalized_content")
         try:
             cursor = db.cursor()
         except Exception as e:
@@ -150,6 +239,8 @@ class ExeSQL(ToolBase, ABC):
                 if pd.api.types.is_datetime64_any_dtype(single_res[col]):
                     single_res[col] = single_res[col].dt.strftime('%Y-%m-%d')
 
+            single_res = single_res.where(pd.notnull(single_res), None)
+
             sql_res.append(convert_decimals(single_res.to_dict(orient='records')))
             formalized_content.append(single_res.to_markdown(index=False, floatfmt=".6f"))
 

agent/tools/pubmed.py

@@ -85,13 +85,7 @@ class PubMed(ToolBase, ABC):
                 self._retrieve_chunks(pubmedcnt.findall("PubmedArticle"),
                                       get_title=lambda child: child.find("MedlineCitation").find("Article").find("ArticleTitle").text,
                                       get_url=lambda child: "https://pubmed.ncbi.nlm.nih.gov/" + child.find("MedlineCitation").find("PMID").text,
-                                      get_content=lambda child: child.find("MedlineCitation") \
+                                      get_content=lambda child: self._format_pubmed_content(child),)
-                                                                    .find("Article") \
-                                                                    .find("Abstract") \
-                                                                    .find("AbstractText").text \
-                                                                    if child.find("MedlineCitation")\
-                                                                            .find("Article").find("Abstract")  \
-                                                                    else "No abstract available")
                 return self.output("formalized_content")
             except Exception as e:
                 last_e = e
@@ -104,5 +98,50 @@ class PubMed(ToolBase, ABC):
 
         assert False, self.output()
 
+    def _format_pubmed_content(self, child):
+        """Extract structured reference info from PubMed XML"""
+        def safe_find(path):
+            node = child
+            for p in path.split("/"):
+                if node is None:
+                    return None
+                node = node.find(p)
+            return node.text if node is not None and node.text else None
+
+        title = safe_find("MedlineCitation/Article/ArticleTitle") or "No title"
+        abstract = safe_find("MedlineCitation/Article/Abstract/AbstractText") or "No abstract available"
+        journal = safe_find("MedlineCitation/Article/Journal/Title") or "Unknown Journal"
+        volume = safe_find("MedlineCitation/Article/Journal/JournalIssue/Volume") or "-"
+        issue = safe_find("MedlineCitation/Article/Journal/JournalIssue/Issue") or "-"
+        pages = safe_find("MedlineCitation/Article/Pagination/MedlinePgn") or "-"
+
+        # Authors
+        authors = []
+        for author in child.findall(".//AuthorList/Author"):
+            lastname = safe_find("LastName") or ""
+            forename = safe_find("ForeName") or ""
+            fullname = f"{forename} {lastname}".strip()
+            if fullname:
+                authors.append(fullname)
+        authors_str = ", ".join(authors) if authors else "Unknown Authors"
+
+        # DOI
+        doi = None
+        for eid in child.findall(".//ArticleId"):
+            if eid.attrib.get("IdType") == "doi":
+                doi = eid.text
+                break
+
+        return (
+            f"Title: {title}\n"
+            f"Authors: {authors_str}\n"
+            f"Journal: {journal}\n"
+            f"Volume: {volume}\n"
+            f"Issue: {issue}\n"
+            f"Pages: {pages}\n"
+            f"DOI: {doi or '-'}\n"
+            f"Abstract: {abstract.strip()}"
+        )
+
     def thoughts(self) -> str:
         return "Looking for scholarly papers on `{}`,” prioritising reputable sources.".format(self.get_input().get("query", "-_-!"))

agent/tools/retrieval.py

@@ -18,12 +18,14 @@ import re
 from abc import ABC
 from agent.tools.base import ToolParamBase, ToolBase, ToolMeta
 from api.db import LLMType
+from api.db.services.document_service import DocumentService
+from api.db.services.dialog_service import meta_filter
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.services.llm_service import LLMBundle
 from api import settings
 from api.utils.api_utils import timeout
 from rag.app.tag import label_question
-from rag.prompts.generator import cross_languages, kb_prompt
+from rag.prompts.generator import cross_languages, kb_prompt, gen_meta_filter
 
 
 class RetrievalParam(ToolParamBase):
@@ -57,6 +59,8 @@ class RetrievalParam(ToolParamBase):
         self.empty_response = ""
         self.use_kg = False
         self.cross_languages = []
+        self.toc_enhance = False
+        self.meta_data_filter={}
 
     def check(self):
         self.check_decimal_float(self.similarity_threshold, "[Retrieval] Similarity threshold")
@@ -116,12 +120,27 @@ class Retrieval(ToolBase, ABC):
         vars = self.get_input_elements_from_text(kwargs["query"])
         vars = {k:o["value"] for k,o in vars.items()}
         query = self.string_format(kwargs["query"], vars)
+        
+        doc_ids=[]
+        if self._param.meta_data_filter!={}:
+            metas = DocumentService.get_meta_by_kbs(kb_ids)
+            if self._param.meta_data_filter.get("method") == "auto":
+                chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT)
+                filters = gen_meta_filter(chat_mdl, metas, query)
+                doc_ids.extend(meta_filter(metas, filters))
+                if not doc_ids:
+                    doc_ids = None
+            elif self._param.meta_data_filter.get("method") == "manual":
+                doc_ids.extend(meta_filter(metas, self._param.meta_data_filter["manual"]))
+                if not doc_ids:
+                    doc_ids = None
+
         if self._param.cross_languages:
             query = cross_languages(kbs[0].tenant_id, None, query, self._param.cross_languages)
 
         if kbs:
             query = re.sub(r"^user[:：\s]*", "", query, flags=re.IGNORECASE)
-            kbinfos = settings.retrievaler.retrieval(
+            kbinfos = settings.retriever.retrieval(
                 query,
                 embd_mdl,
                 [kb.tenant_id for kb in kbs],
@@ -130,12 +149,18 @@ class Retrieval(ToolBase, ABC):
                 self._param.top_n,
                 self._param.similarity_threshold,
                 1 - self._param.keywords_similarity_weight,
+                doc_ids=doc_ids,
                 aggs=False,
                 rerank_mdl=rerank_mdl,
                 rank_feature=label_question(query, kbs),
             )
+            if self._param.toc_enhance:
+                chat_mdl = LLMBundle(self._canvas._tenant_id, LLMType.CHAT)
+                cks = settings.retriever.retrieval_by_toc(query, kbinfos["chunks"], [kb.tenant_id for kb in kbs], chat_mdl, self._param.top_n)
+                if cks:
+                    kbinfos["chunks"] = cks
             if self._param.use_kg:
-                ck = settings.kg_retrievaler.retrieval(query,
+                ck = settings.kg_retriever.retrieval(query,
                                                        [kb.tenant_id for kb in kbs],
                                                        kb_ids,
                                                        embd_mdl,
@@ -146,7 +171,7 @@ class Retrieval(ToolBase, ABC):
             kbinfos = {"chunks": [], "doc_aggs": []}
 
         if self._param.use_kg and kbs:
-            ck = settings.kg_retrievaler.retrieval(query, [kb.tenant_id for kb in kbs], filtered_kb_ids, embd_mdl, LLMBundle(kbs[0].tenant_id, LLMType.CHAT))
+            ck = settings.kg_retriever.retrieval(query, [kb.tenant_id for kb in kbs], filtered_kb_ids, embd_mdl, LLMBundle(kbs[0].tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ck["content"] = ck["content_with_weight"]
                 del ck["content_with_weight"]

agent/tools/searxng.py

@@ -85,7 +85,7 @@ class SearXNG(ToolBase, ABC):
             self.set_output("formalized_content", "")
             return ""
 
-        searxng_url = (kwargs.get("searxng_url") or getattr(self._param, "searxng_url", "") or "").strip()
+        searxng_url = (getattr(self._param, "searxng_url", "") or kwargs.get("searxng_url") or "").strip()
         # In try-run, if no URL configured, just return empty instead of raising
         if not searxng_url:
             self.set_output("formalized_content", "")

api/apps/api_app.py

@@ -536,7 +536,7 @@ def list_chunks():
             )
         kb_ids = KnowledgebaseService.get_kb_ids(tenant_id)
 
-        res = settings.retrievaler.chunk_list(doc_id, tenant_id, kb_ids)
+        res = settings.retriever.chunk_list(doc_id, tenant_id, kb_ids)
         res = [
             {
                 "content": res_item["content_with_weight"],
@@ -884,7 +884,7 @@ def retrieval():
         if req.get("keyword", False):
             chat_mdl = LLMBundle(kbs[0].tenant_id, LLMType.CHAT)
             question += keyword_extraction(chat_mdl, question)
-        ranks = settings.retrievaler.retrieval(question, embd_mdl, kbs[0].tenant_id, kb_ids, page, size,
+        ranks = settings.retriever.retrieval(question, embd_mdl, kbs[0].tenant_id, kb_ids, page, size,
                                                similarity_threshold, vector_similarity_weight, top,
                                                doc_ids, rerank_mdl=rerank_mdl, highlight= highlight,
                                                rank_feature=label_question(question, kbs))

api/apps/canvas_app.py

@@ -19,15 +19,19 @@ import re
 import sys
 from functools import partial
 
+import flask
 import trio
 from flask import request, Response
 from flask_login import login_required, current_user
 
-from agent.component.llm import LLM
+from agent.component import LLM
+from api import settings
 from api.db import CanvasCategory, FileType
 from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService, API4ConversationService
 from api.db.services.document_service import DocumentService
 from api.db.services.file_service import FileService
+from api.db.services.pipeline_operation_log_service import PipelineOperationLogService
+from api.db.services.task_service import queue_dataflow, CANVAS_DEBUG_DOC_ID, TaskService
 from api.db.services.user_service import TenantService
 from api.db.services.user_canvas_version import UserCanvasVersionService
 from api.settings import RetCode
@@ -35,25 +39,19 @@ from api.utils import get_uuid
 from api.utils.api_utils import get_json_result, server_error_response, validate_request, get_data_error_result
 from agent.canvas import Canvas
 from peewee import MySQLDatabase, PostgresqlDatabase
-from api.db.db_models import APIToken
+from api.db.db_models import APIToken, Task
 import time
 
 from api.utils.file_utils import filename_type, read_potential_broken_pdf
+from rag.flow.pipeline import Pipeline
+from rag.nlp import search
 from rag.utils.redis_conn import REDIS_CONN
 
 
 @manager.route('/templates', methods=['GET'])  # noqa: F821
 @login_required
 def templates():
-    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.Agent)])
+    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.get_all()])
-
-
-@manager.route('/list', methods=['GET'])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in \
-                                 UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.Agent)], key=lambda x: x["update_time"]*-1)
-                           )
 
 
 @manager.route('/rm', methods=['POST'])  # noqa: F821
@@ -77,9 +75,10 @@ def save():
     if not isinstance(req["dsl"], str):
         req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
     req["dsl"] = json.loads(req["dsl"])
+    cate = req.get("canvas_category", CanvasCategory.Agent)
     if "id" not in req:
         req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.Agent):
+        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=cate):
             return get_data_error_result(message=f"{req['title'].strip()} already exists.")
         req["id"] = get_uuid()
         if not UserCanvasService.save(**req):
@@ -101,7 +100,7 @@ def save():
 def get(canvas_id):
     if not UserCanvasService.accessible(canvas_id, current_user.id):
         return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, c = UserCanvasService.get_by_canvas_id(canvas_id)
     return get_json_result(data=c)
 
 
@@ -148,6 +147,14 @@ def run():
     if not isinstance(cvs.dsl, str):
         cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
 
+    if cvs.canvas_category == CanvasCategory.DataFlow:
+        task_id = get_uuid()
+        Pipeline(cvs.dsl, tenant_id=current_user.id, doc_id=CANVAS_DEBUG_DOC_ID, task_id=task_id, flow_id=req["id"])
+        ok, error_message = queue_dataflow(tenant_id=user_id, flow_id=req["id"], task_id=task_id, file=files[0], priority=0)
+        if not ok:
+            return get_data_error_result(message=error_message)
+        return get_json_result(data={"message_id": task_id})
+
     try:
         canvas = Canvas(cvs.dsl, current_user.id, req["id"])
     except Exception as e:
@@ -173,6 +180,44 @@ def run():
     return resp
 
 
+@manager.route('/rerun', methods=['POST'])  # noqa: F821
+@validate_request("id", "dsl", "component_id")
+@login_required
+def rerun():
+    req = request.json
+    doc = PipelineOperationLogService.get_documents_info(req["id"])
+    if not doc:
+        return get_data_error_result(message="Document not found.")
+    doc = doc[0]
+    if 0 < doc["progress"] < 1:
+        return get_data_error_result(message=f"`{doc['name']}` is processing...")
+
+    if settings.docStoreConn.indexExist(search.index_name(current_user.id), doc["kb_id"]):
+        settings.docStoreConn.delete({"doc_id": doc["id"]}, search.index_name(current_user.id), doc["kb_id"])
+    doc["progress_msg"] = ""
+    doc["chunk_num"] = 0
+    doc["token_num"] = 0
+    DocumentService.clear_chunk_num_when_rerun(doc["id"])
+    DocumentService.update_by_id(id, doc)
+    TaskService.filter_delete([Task.doc_id == id])
+
+    dsl = req["dsl"]
+    dsl["path"] = [req["component_id"]]
+    PipelineOperationLogService.update_by_id(req["id"], {"dsl": dsl})
+    queue_dataflow(tenant_id=current_user.id, flow_id=req["id"], task_id=get_uuid(), doc_id=doc["id"], priority=0, rerun=True)
+    return get_json_result(data=True)
+
+
+@manager.route('/cancel/<task_id>', methods=['PUT'])  # noqa: F821
+@login_required
+def cancel(task_id):
+    try:
+        REDIS_CONN.set(f"{task_id}-cancel", "x")
+    except Exception as e:
+        logging.exception(e)
+    return get_json_result(data=True)
+
+
 @manager.route('/reset', methods=['POST'])  # noqa: F821
 @validate_request("id")
 @login_required
@@ -198,7 +243,7 @@ def reset():
 
 @manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
 def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_id(canvas_id)
+    e, cvs = UserCanvasService.get_by_canvas_id(canvas_id)
     if not e:
         return get_data_error_result(message="canvas not found.")
 
@@ -348,6 +393,65 @@ def test_db_connect():
             cursor = db.cursor()
             cursor.execute("SELECT 1")
             cursor.close()
+        elif req["db_type"] == 'IBM DB2':
+            import ibm_db
+            conn_str = (
+                f"DATABASE={req['database']};"
+                f"HOSTNAME={req['host']};"
+                f"PORT={req['port']};"
+                f"PROTOCOL=TCPIP;"
+                f"UID={req['username']};"
+                f"PWD={req['password']};"
+            )
+            logging.info(conn_str)
+            conn = ibm_db.connect(conn_str, "", "")
+            stmt = ibm_db.exec_immediate(conn, "SELECT 1 FROM sysibm.sysdummy1")
+            ibm_db.fetch_assoc(stmt)
+            ibm_db.close(conn)
+            return get_json_result(data="Database Connection Successful!")
+        elif req["db_type"] == 'trino':
+            def _parse_catalog_schema(db: str):
+                if not db:
+                    return None, None
+                if "." in db:
+                    c, s = db.split(".", 1)
+                elif "/" in db:
+                    c, s = db.split("/", 1)
+                else:
+                    c, s = db, "default"
+                return c, s
+            try:
+                import trino
+                import os
+                from trino.auth import BasicAuthentication
+            except Exception:
+                return server_error_response("Missing dependency 'trino'. Please install: pip install trino")
+
+            catalog, schema = _parse_catalog_schema(req["database"])
+            if not catalog:
+                return server_error_response("For Trino, 'database' must be 'catalog.schema' or at least 'catalog'.")
+            
+            http_scheme = "https" if os.environ.get("TRINO_USE_TLS", "0") == "1" else "http"
+
+            auth = None
+            if http_scheme == "https" and req.get("password"):
+                auth = BasicAuthentication(req.get("username") or "ragflow", req["password"])
+
+            conn = trino.dbapi.connect(
+                host=req["host"],
+                port=int(req["port"] or 8080),
+                user=req["username"] or "ragflow",
+                catalog=catalog,
+                schema=schema or "default",
+                http_scheme=http_scheme,
+                auth=auth
+            )
+            cur = conn.cursor()
+            cur.execute("SELECT 1")
+            cur.fetchall()
+            cur.close()
+            conn.close()
+            return get_json_result(data="Database Connection Successful!")
         else:
             return server_error_response("Unsupported database type.")
         if req["db_type"] != 'mssql':
@@ -383,22 +487,32 @@ def getversion( version_id):
         return get_json_result(data=f"Error getting history file: {e}")
 
 
-@manager.route('/listteam', methods=['GET'])  # noqa: F821
+@manager.route('/list', methods=['GET'])  # noqa: F821
 @login_required
 def list_canvas():
     keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 1))
+    page_number = int(request.args.get("page", 0))
-    items_per_page = int(request.args.get("page_size", 150))
+    items_per_page = int(request.args.get("page_size", 0))
     orderby = request.args.get("orderby", "create_time")
-    desc = request.args.get("desc", True)
+    canvas_category = request.args.get("canvas_category")
-    try:
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    owner_ids = [id for id in request.args.get("owner_ids", "").strip().split(",") if id]
+    if not owner_ids:
         tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
+        tenants = [m["tenant_id"] for m in tenants]
+        tenants.append(current_user.id)
         canvas, total = UserCanvasService.get_by_tenant_ids(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number,
+            tenants, current_user.id, page_number,
-            items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.Agent)
+            items_per_page, orderby, desc, keywords, canvas_category)
-        return get_json_result(data={"canvas": canvas, "total": total})
+    else:
-    except Exception as e:
+        tenants = owner_ids
-        return server_error_response(e)
+        canvas, total = UserCanvasService.get_by_tenant_ids(
+            tenants, current_user.id, 0,
+            0, orderby, desc, keywords, canvas_category)
+    return get_json_result(data={"canvas": canvas, "total": total})
 
 
 @manager.route('/setting', methods=['POST'])  # noqa: F821
@@ -483,3 +597,11 @@ def prompts():
         #"context_ranking": RANK_MEMORY,
         "citation_guidelines": CITATION_PROMPT_TEMPLATE
     })
+
+
+@manager.route('/download', methods=['GET'])  # noqa: F821
+def download():
+    id = request.args.get("id")
+    created_by = request.args.get("created_by")
+    blob = FileService.get_blob(created_by, id)
+    return flask.make_response(blob)

api/apps/chunk_app.py

@@ -60,7 +60,7 @@ def list_chunk():
         }
         if "available_int" in req:
             query["available_int"] = int(req["available_int"])
-        sres = settings.retrievaler.search(query, search.index_name(tenant_id), kb_ids, highlight=True)
+        sres = settings.retriever.search(query, search.index_name(tenant_id), kb_ids, highlight=["content_ltks"])
         res = {"total": sres.total, "chunks": [], "doc": doc.to_dict()}
         for id in sres.ids:
             d = {
@@ -346,15 +346,16 @@ def retrieval_test():
             question += keyword_extraction(chat_mdl, question)
 
         labels = label_question(question, [kb])
-        ranks = settings.retrievaler.retrieval(question, embd_mdl, tenant_ids, kb_ids, page, size,
+        ranks = settings.retriever.retrieval(question, embd_mdl, tenant_ids, kb_ids, page, size,
                                float(req.get("similarity_threshold", 0.0)),
                                float(req.get("vector_similarity_weight", 0.3)),
                                top,
-                               doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"),
+                               doc_ids, rerank_mdl=rerank_mdl,
+                                             highlight=req.get("highlight", False),
                                rank_feature=labels
                                )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question,
+            ck = settings.kg_retriever.retrieval(question,
                                                    tenant_ids,
                                                    kb_ids,
                                                    embd_mdl,
@@ -384,7 +385,7 @@ def knowledge_graph():
         "doc_ids": [doc_id],
         "knowledge_graph_kwd": ["graph", "mind_map"]
     }
-    sres = settings.retrievaler.search(req, search.index_name(tenant_id), kb_ids)
+    sres = settings.retriever.search(req, search.index_name(tenant_id), kb_ids)
     obj = {"graph": {}, "mind_map": {}}
     for id in sres.ids[:2]:
         ty = sres.field[id]["knowledge_graph_kwd"]

api/apps/dataflow_app.py

@@ -1,353 +0,0 @@
-#
-#  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
-#
-#  Licensed under the Apache License, Version 2.0 (the "License");
-#  you may not use this file except in compliance with the License.
-#  You may obtain a copy of the License at
-#
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-#  Unless required by applicable law or agreed to in writing, software
-#  distributed under the License is distributed on an "AS IS" BASIS,
-#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-#  See the License for the specific language governing permissions and
-#  limitations under the License.
-#
-import json
-import re
-import sys
-import time
-from functools import partial
-
-import trio
-from flask import request
-from flask_login import current_user, login_required
-
-from agent.canvas import Canvas
-from agent.component.llm import LLM
-from api.db import CanvasCategory, FileType
-from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService
-from api.db.services.document_service import DocumentService
-from api.db.services.file_service import FileService
-from api.db.services.task_service import queue_dataflow
-from api.db.services.user_canvas_version import UserCanvasVersionService
-from api.db.services.user_service import TenantService
-from api.settings import RetCode
-from api.utils import get_uuid
-from api.utils.api_utils import get_data_error_result, get_json_result, server_error_response, validate_request
-from api.utils.file_utils import filename_type, read_potential_broken_pdf
-from rag.flow.pipeline import Pipeline
-
-
-@manager.route("/templates", methods=["GET"])  # noqa: F821
-@login_required
-def templates():
-    return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.DataFlow)])
-
-
-@manager.route("/list", methods=["GET"])  # noqa: F821
-@login_required
-def canvas_list():
-    return get_json_result(data=sorted([c.to_dict() for c in UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.DataFlow)], key=lambda x: x["update_time"] * -1))
-
-
-@manager.route("/rm", methods=["POST"])  # noqa: F821
-@validate_request("canvas_ids")
-@login_required
-def rm():
-    for i in request.json["canvas_ids"]:
-        if not UserCanvasService.accessible(i, current_user.id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-        UserCanvasService.delete_by_id(i)
-    return get_json_result(data=True)
-
-
-@manager.route("/set", methods=["POST"])  # noqa: F821
-@validate_request("dsl", "title")
-@login_required
-def save():
-    req = request.json
-    if not isinstance(req["dsl"], str):
-        req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
-    req["dsl"] = json.loads(req["dsl"])
-    req["canvas_category"] = CanvasCategory.DataFlow
-    if "id" not in req:
-        req["user_id"] = current_user.id
-        if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.DataFlow):
-            return get_data_error_result(message=f"{req['title'].strip()} already exists.")
-        req["id"] = get_uuid()
-
-        if not UserCanvasService.save(**req):
-            return get_data_error_result(message="Fail to save canvas.")
-    else:
-        if not UserCanvasService.accessible(req["id"], current_user.id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-        UserCanvasService.update_by_id(req["id"], req)
-    # save version
-    UserCanvasVersionService.insert(user_canvas_id=req["id"], dsl=req["dsl"], title="{0}_{1}".format(req["title"], time.strftime("%Y_%m_%d_%H_%M_%S")))
-    UserCanvasVersionService.delete_all_versions(req["id"])
-    return get_json_result(data=req)
-
-
-@manager.route("/get/<canvas_id>", methods=["GET"])  # noqa: F821
-@login_required
-def get(canvas_id):
-    if not UserCanvasService.accessible(canvas_id, current_user.id):
-        return get_data_error_result(message="canvas not found.")
-    e, c = UserCanvasService.get_by_tenant_id(canvas_id)
-    return get_json_result(data=c)
-
-
-@manager.route("/run", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def run():
-    req = request.json
-    flow_id = req.get("id", "")
-    doc_id = req.get("doc_id", "")
-    if not all([flow_id, doc_id]):
-        return get_data_error_result(message="id and doc_id are required.")
-
-    if not DocumentService.get_by_id(doc_id):
-        return get_data_error_result(message=f"Document for {doc_id} not found.")
-
-    user_id = req.get("user_id", current_user.id)
-    if not UserCanvasService.accessible(flow_id, current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    e, cvs = UserCanvasService.get_by_id(flow_id)
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-
-    if not isinstance(cvs.dsl, str):
-        cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
-
-    task_id = get_uuid()
-
-    ok, error_message = queue_dataflow(dsl=cvs.dsl, tenant_id=user_id, doc_id=doc_id, task_id=task_id, flow_id=flow_id, priority=0)
-    if not ok:
-        return server_error_response(error_message)
-
-    return get_json_result(data={"task_id": task_id, "flow_id": flow_id})
-
-
-@manager.route("/reset", methods=["POST"])  # noqa: F821
-@validate_request("id")
-@login_required
-def reset():
-    req = request.json
-    flow_id = req.get("id", "")
-    if not flow_id:
-        return get_data_error_result(message="id is required.")
-
-    if not UserCanvasService.accessible(flow_id, current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    task_id = req.get("task_id", "")
-
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id=task_id)
-        dataflow.reset()
-        req["dsl"] = json.loads(str(dataflow))
-        UserCanvasService.update_by_id(req["id"], {"dsl": req["dsl"]})
-        return get_json_result(data=req["dsl"])
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/upload/<canvas_id>", methods=["POST"])  # noqa: F821
-def upload(canvas_id):
-    e, cvs = UserCanvasService.get_by_tenant_id(canvas_id)
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-
-    user_id = cvs["user_id"]
-
-    def structured(filename, filetype, blob, content_type):
-        nonlocal user_id
-        if filetype == FileType.PDF.value:
-            blob = read_potential_broken_pdf(blob)
-
-        location = get_uuid()
-        FileService.put_blob(user_id, location, blob)
-
-        return {
-            "id": location,
-            "name": filename,
-            "size": sys.getsizeof(blob),
-            "extension": filename.split(".")[-1].lower(),
-            "mime_type": content_type,
-            "created_by": user_id,
-            "created_at": time.time(),
-            "preview_url": None,
-        }
-
-    if request.args.get("url"):
-        from crawl4ai import AsyncWebCrawler, BrowserConfig, CrawlerRunConfig, CrawlResult, DefaultMarkdownGenerator, PruningContentFilter
-
-        try:
-            url = request.args.get("url")
-            filename = re.sub(r"\?.*", "", url.split("/")[-1])
-
-            async def adownload():
-                browser_config = BrowserConfig(
-                    headless=True,
-                    verbose=False,
-                )
-                async with AsyncWebCrawler(config=browser_config) as crawler:
-                    crawler_config = CrawlerRunConfig(markdown_generator=DefaultMarkdownGenerator(content_filter=PruningContentFilter()), pdf=True, screenshot=False)
-                    result: CrawlResult = await crawler.arun(url=url, config=crawler_config)
-                    return result
-
-            page = trio.run(adownload())
-            if page.pdf:
-                if filename.split(".")[-1].lower() != "pdf":
-                    filename += ".pdf"
-                return get_json_result(data=structured(filename, "pdf", page.pdf, page.response_headers["content-type"]))
-
-            return get_json_result(data=structured(filename, "html", str(page.markdown).encode("utf-8"), page.response_headers["content-type"], user_id))
-
-        except Exception as e:
-            return server_error_response(e)
-
-    file = request.files["file"]
-    try:
-        DocumentService.check_doc_health(user_id, file.filename)
-        return get_json_result(data=structured(file.filename, filename_type(file.filename), file.read(), file.content_type))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/input_form", methods=["GET"])  # noqa: F821
-@login_required
-def input_form():
-    flow_id = request.args.get("id")
-    cpn_id = request.args.get("component_id")
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(flow_id)
-        if not e:
-            return get_data_error_result(message="canvas not found.")
-        if not UserCanvasService.query(user_id=current_user.id, id=flow_id):
-            return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-        dataflow = Pipeline(dsl=json.dumps(user_canvas.dsl), tenant_id=current_user.id, flow_id=flow_id, task_id="")
-
-        return get_json_result(data=dataflow.get_component_input_form(cpn_id))
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/debug", methods=["POST"])  # noqa: F821
-@validate_request("id", "component_id", "params")
-@login_required
-def debug():
-    req = request.json
-    if not UserCanvasService.accessible(req["id"], current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-    try:
-        e, user_canvas = UserCanvasService.get_by_id(req["id"])
-        canvas = Canvas(json.dumps(user_canvas.dsl), current_user.id)
-        canvas.reset()
-        canvas.message_id = get_uuid()
-        component = canvas.get_component(req["component_id"])["obj"]
-        component.reset()
-
-        if isinstance(component, LLM):
-            component.set_debug_inputs(req["params"])
-        component.invoke(**{k: o["value"] for k, o in req["params"].items()})
-        outputs = component.output()
-        for k in outputs.keys():
-            if isinstance(outputs[k], partial):
-                txt = ""
-                for c in outputs[k]():
-                    txt += c
-                outputs[k] = txt
-        return get_json_result(data=outputs)
-    except Exception as e:
-        return server_error_response(e)
-
-
-# api get list version dsl of canvas
-@manager.route("/getlistversion/<canvas_id>", methods=["GET"])  # noqa: F821
-@login_required
-def getlistversion(canvas_id):
-    try:
-        list = sorted([c.to_dict() for c in UserCanvasVersionService.list_by_canvas_id(canvas_id)], key=lambda x: x["update_time"] * -1)
-        return get_json_result(data=list)
-    except Exception as e:
-        return get_data_error_result(message=f"Error getting history files: {e}")
-
-
-# api get version dsl of canvas
-@manager.route("/getversion/<version_id>", methods=["GET"])  # noqa: F821
-@login_required
-def getversion(version_id):
-    try:
-        e, version = UserCanvasVersionService.get_by_id(version_id)
-        if version:
-            return get_json_result(data=version.to_dict())
-    except Exception as e:
-        return get_json_result(data=f"Error getting history file: {e}")
-
-
-@manager.route("/listteam", methods=["GET"])  # noqa: F821
-@login_required
-def list_canvas():
-    keywords = request.args.get("keywords", "")
-    page_number = int(request.args.get("page", 1))
-    items_per_page = int(request.args.get("page_size", 150))
-    orderby = request.args.get("orderby", "create_time")
-    desc = request.args.get("desc", True)
-    try:
-        tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
-        canvas, total = UserCanvasService.get_by_tenant_ids(
-            [m["tenant_id"] for m in tenants], current_user.id, page_number, items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.DataFlow
-        )
-        return get_json_result(data={"canvas": canvas, "total": total})
-    except Exception as e:
-        return server_error_response(e)
-
-
-@manager.route("/setting", methods=["POST"])  # noqa: F821
-@validate_request("id", "title", "permission")
-@login_required
-def setting():
-    req = request.json
-    req["user_id"] = current_user.id
-
-    if not UserCanvasService.accessible(req["id"], current_user.id):
-        return get_json_result(data=False, message="Only owner of canvas authorized for this operation.", code=RetCode.OPERATING_ERROR)
-
-    e, flow = UserCanvasService.get_by_id(req["id"])
-    if not e:
-        return get_data_error_result(message="canvas not found.")
-    flow = flow.to_dict()
-    flow["title"] = req["title"]
-    for key in ("description", "permission", "avatar"):
-        if value := req.get(key):
-            flow[key] = value
-
-    num = UserCanvasService.update_by_id(req["id"], flow)
-    return get_json_result(data=num)
-
-
-@manager.route("/trace", methods=["GET"])  # noqa: F821
-def trace():
-    dataflow_id = request.args.get("dataflow_id")
-    task_id = request.args.get("task_id")
-    if not all([dataflow_id, task_id]):
-        return get_data_error_result(message="dataflow_id and task_id are required.")
-
-    e, dataflow_canvas = UserCanvasService.get_by_id(dataflow_id)
-    if not e:
-        return get_data_error_result(message="dataflow not found.")
-
-    dsl_str = json.dumps(dataflow_canvas.dsl, ensure_ascii=False)
-    dataflow = Pipeline(dsl=dsl_str, tenant_id=dataflow_canvas.user_id, flow_id=dataflow_id, task_id=task_id)
-    log = dataflow.fetch_logs()
-
-    return get_json_result(data=log)

api/apps/document_app.py

@@ -24,6 +24,7 @@ from flask import request
 from flask_login import current_user, login_required
 
 from api import settings
+from api.common.check_team_permission import check_kb_team_permission
 from api.constants import FILE_NAME_LEN_LIMIT, IMG_BASE64_PREFIX
 from api.db import VALID_FILE_TYPES, VALID_TASK_STATUS, FileSource, FileType, ParserType, TaskStatus
 from api.db.db_models import File, Task
@@ -32,7 +33,7 @@ from api.db.services.document_service import DocumentService, doc_upload_and_par
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
 from api.db.services.knowledgebase_service import KnowledgebaseService
-from api.db.services.task_service import TaskService, cancel_all_task_of, queue_tasks
+from api.db.services.task_service import TaskService, cancel_all_task_of, queue_tasks, queue_dataflow
 from api.db.services.user_service import UserTenantService
 from api.utils import get_uuid
 from api.utils.api_utils import (
@@ -44,7 +45,7 @@ from api.utils.api_utils import (
 from api.utils.file_utils import filename_type, get_project_base_directory, thumbnail
 from api.utils.web_utils import CONTENT_TYPE_MAP, html2pdf, is_valid_url
 from deepdoc.parser.html_parser import RAGFlowHtmlParser
-from rag.nlp import search
+from rag.nlp import search, rag_tokenizer
 from rag.utils.storage_factory import STORAGE_IMPL
 
 
@@ -68,8 +69,10 @@ def upload():
     e, kb = KnowledgebaseService.get_by_id(kb_id)
     if not e:
         raise LookupError("Can't find this knowledgebase!")
-    err, files = FileService.upload_document(kb, file_objs, current_user.id)
+    if not check_kb_team_permission(kb, current_user.id):
+        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
 
+    err, files = FileService.upload_document(kb, file_objs, current_user.id)
     if err:
         return get_json_result(data=files, message="\n".join(err), code=settings.RetCode.SERVER_ERROR)
 
@@ -94,6 +97,8 @@ def web_crawl():
     e, kb = KnowledgebaseService.get_by_id(kb_id)
     if not e:
         raise LookupError("Can't find this knowledgebase!")
+    if check_kb_team_permission(kb, current_user.id):
+        return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
 
     blob = html2pdf(url)
     if not blob:
@@ -182,6 +187,7 @@ def create():
                 "id": get_uuid(),
                 "kb_id": kb.id,
                 "parser_id": kb.parser_id,
+                "pipeline_id": kb.pipeline_id,
                 "parser_config": kb.parser_config,
                 "created_by": current_user.id,
                 "type": FileType.VIRTUAL,
@@ -479,8 +485,11 @@ def run():
                         kb_table_num_map[kb_id] = count
                         if kb_table_num_map[kb_id] <= 0:
                             KnowledgebaseService.delete_field_map(kb_id)
-                bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
+                if doc.get("pipeline_id", ""):
-                queue_tasks(doc, bucket, name, 0)
+                    queue_dataflow(tenant_id, flow_id=doc["pipeline_id"], task_id=get_uuid(), doc_id=id)
+                else:
+                    bucket, name = File2DocumentService.get_storage_address(doc_id=doc["id"])
+                    queue_tasks(doc, bucket, name, 0)
 
         return get_json_result(data=True)
     except Exception as e:
@@ -515,6 +524,21 @@ def rename():
             e, file = FileService.get_by_id(informs[0].file_id)
             FileService.update_by_id(file.id, {"name": req["name"]})
 
+        tenant_id = DocumentService.get_tenant_id(req["doc_id"])
+        title_tks = rag_tokenizer.tokenize(req["name"])
+        es_body = {
+            "docnm_kwd": req["name"],
+            "title_tks": title_tks,
+            "title_sm_tks": rag_tokenizer.fine_grained_tokenize(title_tks),
+        }
+        if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
+            settings.docStoreConn.update(
+                {"doc_id": req["doc_id"]},
+                es_body,
+                search.index_name(tenant_id),
+                doc.kb_id,
+            )
+
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
@@ -546,31 +570,22 @@ def get(doc_id):
 
 @manager.route("/change_parser", methods=["POST"])  # noqa: F821
 @login_required
-@validate_request("doc_id", "parser_id")
+@validate_request("doc_id")
 def change_parser():
-    req = request.json
 
+    req = request.json
     if not DocumentService.accessible(req["doc_id"], current_user.id):
         return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
-    try:
+
-        e, doc = DocumentService.get_by_id(req["doc_id"])
+    e, doc = DocumentService.get_by_id(req["doc_id"])
+    if not e:
+        return get_data_error_result(message="Document not found!")
+
+    def reset_doc():
+        nonlocal doc
+        e = DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"], "parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
         if not e:
             return get_data_error_result(message="Document not found!")
-        if doc.parser_id.lower() == req["parser_id"].lower():
-            if "parser_config" in req:
-                if req["parser_config"] == doc.parser_config:
-                    return get_json_result(data=True)
-            else:
-                return get_json_result(data=True)
-
-        if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
-            return get_data_error_result(message="Not supported yet!")
-
-        e = DocumentService.update_by_id(doc.id, {"parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
-        if not e:
-            return get_data_error_result(message="Document not found!")
-        if "parser_config" in req:
-            DocumentService.update_parser_config(doc.id, req["parser_config"])
         if doc.token_num > 0:
             e = DocumentService.increment_chunk_num(doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, doc.process_duration * -1)
             if not e:
@@ -581,6 +596,26 @@ def change_parser():
             if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
                 settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
 
+    try:
+        if "pipeline_id" in req and req["pipeline_id"] != "":
+            if doc.pipeline_id == req["pipeline_id"]:
+                return get_json_result(data=True)
+            DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"]})
+            reset_doc()
+            return get_json_result(data=True)
+
+        if doc.parser_id.lower() == req["parser_id"].lower():
+            if "parser_config" in req:
+                if req["parser_config"] == doc.parser_config:
+                    return get_json_result(data=True)
+            else:
+                return get_json_result(data=True)
+
+        if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
+            return get_data_error_result(message="Not supported yet!")
+        if "parser_config" in req:
+            DocumentService.update_parser_config(doc.id, req["parser_config"])
+        reset_doc()
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)

api/apps/file_app.py

@@ -13,6 +13,7 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License
 #
+import logging
 import os
 import pathlib
 import re
@@ -21,6 +22,7 @@ import flask
 from flask import request
 from flask_login import login_required, current_user
 
+from api.common.check_team_permission import check_file_team_permission
 from api.db.services.document_service import DocumentService
 from api.db.services.file2document_service import File2DocumentService
 from api.utils.api_utils import server_error_response, get_data_error_result, validate_request
@@ -233,52 +235,63 @@ def get_all_parent_folders():
         return server_error_response(e)
 
 
-@manager.route('/rm', methods=['POST'])  # noqa: F821
+@manager.route("/rm", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("file_ids")
 def rm():
     req = request.json
     file_ids = req["file_ids"]
+
+    def _delete_single_file(file):
+        try:
+            if file.location:
+                STORAGE_IMPL.rm(file.parent_id, file.location)
+        except Exception:
+            logging.exception(f"Fail to remove object: {file.parent_id}/{file.location}")
+
+        informs = File2DocumentService.get_by_file_id(file.id)
+        for inform in informs:
+            doc_id = inform.document_id
+            e, doc = DocumentService.get_by_id(doc_id)
+            if e and doc:
+                tenant_id = DocumentService.get_tenant_id(doc_id)
+                if tenant_id:
+                    DocumentService.remove_document(doc, tenant_id)
+            File2DocumentService.delete_by_file_id(file.id)
+
+        FileService.delete(file)
+
+    def _delete_folder_recursive(folder, tenant_id):
+        sub_files = FileService.list_all_files_by_parent_id(folder.id)
+        for sub_file in sub_files:
+            if sub_file.type == FileType.FOLDER.value:
+                _delete_folder_recursive(sub_file, tenant_id)
+            else:
+                _delete_single_file(sub_file)
+
+        FileService.delete(folder)
+
     try:
         for file_id in file_ids:
             e, file = FileService.get_by_id(file_id)
-            if not e:
+            if not e or not file:
                 return get_data_error_result(message="File or Folder not found!")
             if not file.tenant_id:
                 return get_data_error_result(message="Tenant not found!")
+            if not check_file_team_permission(file, current_user.id):
+                return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
+
             if file.source_type == FileSource.KNOWLEDGEBASE:
                 continue
 
             if file.type == FileType.FOLDER.value:
-                file_id_list = FileService.get_all_innermost_file_ids(file_id, [])
+                _delete_folder_recursive(file, current_user.id)
-                for inner_file_id in file_id_list:
+                continue
-                    e, file = FileService.get_by_id(inner_file_id)
-                    if not e:
-                        return get_data_error_result(message="File not found!")
-                    STORAGE_IMPL.rm(file.parent_id, file.location)
-                FileService.delete_folder_by_pf_id(current_user.id, file_id)
-            else:
-                STORAGE_IMPL.rm(file.parent_id, file.location)
-                if not FileService.delete(file):
-                    return get_data_error_result(
-                        message="Database error (File removal)!")
 
-            # delete file2document
+            _delete_single_file(file)
-            informs = File2DocumentService.get_by_file_id(file_id)
-            for inform in informs:
-                doc_id = inform.document_id
-                e, doc = DocumentService.get_by_id(doc_id)
-                if not e:
-                    return get_data_error_result(message="Document not found!")
-                tenant_id = DocumentService.get_tenant_id(doc_id)
-                if not tenant_id:
-                    return get_data_error_result(message="Tenant not found!")
-                if not DocumentService.remove_document(doc, tenant_id):
-                    return get_data_error_result(
-                        message="Database error (Document removal)!")
-            File2DocumentService.delete_by_file_id(file_id)
 
         return get_json_result(data=True)
+
     except Exception as e:
         return server_error_response(e)
 
@@ -292,6 +305,8 @@ def rename():
         e, file = FileService.get_by_id(req["file_id"])
         if not e:
             return get_data_error_result(message="File not found!")
+        if not check_file_team_permission(file, current_user.id):
+            return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
         if file.type != FileType.FOLDER.value \
             and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(
                 file.name.lower()).suffix:
@@ -328,6 +343,8 @@ def get(file_id):
         e, file = FileService.get_by_id(file_id)
         if not e:
             return get_data_error_result(message="Document not found!")
+        if not check_file_team_permission(file, current_user.id):
+            return get_json_result(data=False, message='No authorization.', code=settings.RetCode.AUTHENTICATION_ERROR)
 
         blob = STORAGE_IMPL.get(file.parent_id, file.location)
         if not blob:
@@ -348,29 +365,89 @@ def get(file_id):
         return server_error_response(e)
 
 
-@manager.route('/mv', methods=['POST'])  # noqa: F821
+@manager.route("/mv", methods=["POST"])  # noqa: F821
 @login_required
 @validate_request("src_file_ids", "dest_file_id")
 def move():
     req = request.json
     try:
         file_ids = req["src_file_ids"]
-        parent_id = req["dest_file_id"]
+        dest_parent_id = req["dest_file_id"]
+
+        ok, dest_folder = FileService.get_by_id(dest_parent_id)
+        if not ok or not dest_folder:
+            return get_data_error_result(message="Parent Folder not found!")
+
         files = FileService.get_by_ids(file_ids)
-        files_dict = {}
+        if not files:
-        for file in files:
+            return get_data_error_result(message="Source files not found!")
-            files_dict[file.id] = file
+
+        files_dict = {f.id: f for f in files}
 
         for file_id in file_ids:
-            file = files_dict[file_id]
+            file = files_dict.get(file_id)
             if not file:
                 return get_data_error_result(message="File or Folder not found!")
             if not file.tenant_id:
                 return get_data_error_result(message="Tenant not found!")
-        fe, _ = FileService.get_by_id(parent_id)
+            if not check_file_team_permission(file, current_user.id):
-        if not fe:
+                return get_json_result(
-            return get_data_error_result(message="Parent Folder not found!")
+                    data=False,
-        FileService.move_file(file_ids, parent_id)
+                    message="No authorization.",
+                    code=settings.RetCode.AUTHENTICATION_ERROR,
+                )
+
+        def _move_entry_recursive(source_file_entry, dest_folder):
+            if source_file_entry.type == FileType.FOLDER.value:
+                existing_folder = FileService.query(name=source_file_entry.name, parent_id=dest_folder.id)
+                if existing_folder:
+                    new_folder = existing_folder[0]
+                else:
+                    new_folder = FileService.insert(
+                        {
+                            "id": get_uuid(),
+                            "parent_id": dest_folder.id,
+                            "tenant_id": source_file_entry.tenant_id,
+                            "created_by": current_user.id,
+                            "name": source_file_entry.name,
+                            "location": "",
+                            "size": 0,
+                            "type": FileType.FOLDER.value,
+                        }
+                    )
+
+                sub_files = FileService.list_all_files_by_parent_id(source_file_entry.id)
+                for sub_file in sub_files:
+                    _move_entry_recursive(sub_file, new_folder)
+
+                FileService.delete_by_id(source_file_entry.id)
+                return
+
+            old_parent_id = source_file_entry.parent_id
+            old_location = source_file_entry.location
+            filename = source_file_entry.name
+
+            new_location = filename
+            while STORAGE_IMPL.obj_exist(dest_folder.id, new_location):
+                new_location += "_"
+
+            try:
+                STORAGE_IMPL.move(old_parent_id, old_location, dest_folder.id, new_location)
+            except Exception as storage_err:
+                raise RuntimeError(f"Move file failed at storage layer: {str(storage_err)}")
+
+            FileService.update_by_id(
+                source_file_entry.id,
+                {
+                    "parent_id": dest_folder.id,
+                    "location": new_location,
+                },
+            )
+
+        for file in files:
+            _move_entry_recursive(file, dest_folder)
+
         return get_json_result(data=True)
+
     except Exception as e:
         return server_error_response(e)

api/apps/kb_app.py

@@ -14,18 +14,21 @@
 #  limitations under the License.
 #
 import json
+import logging
 
 from flask import request
 from flask_login import login_required, current_user
 
 from api.db.services import duplicate_name
-from api.db.services.document_service import DocumentService
+from api.db.services.document_service import DocumentService, queue_raptor_o_graphrag_tasks
 from api.db.services.file2document_service import File2DocumentService
 from api.db.services.file_service import FileService
+from api.db.services.pipeline_operation_log_service import PipelineOperationLogService
+from api.db.services.task_service import TaskService, GRAPH_RAPTOR_FAKE_DOC_ID
 from api.db.services.user_service import TenantService, UserTenantService
-from api.utils.api_utils import server_error_response, get_data_error_result, validate_request, not_allowed_parameters, active_required
+from api.utils.api_utils import get_error_data_result, server_error_response, get_data_error_result, validate_request, not_allowed_parameters
 from api.utils import get_uuid
-from api.db import StatusEnum, FileSource
+from api.db import PipelineTaskType, StatusEnum, FileSource, VALID_FILE_TYPES, VALID_TASK_STATUS
 from api.db.services.knowledgebase_service import KnowledgebaseService
 from api.db.db_models import File
 from api.utils.api_utils import get_json_result
@@ -33,12 +36,12 @@ from api import settings
 from rag.nlp import search
 from api.constants import DATASET_NAME_LIMIT
 from rag.settings import PAGERANK_FLD
+from rag.utils.redis_conn import REDIS_CONN
 from rag.utils.storage_factory import STORAGE_IMPL
 
 
 @manager.route('/create', methods=['post'])  # noqa: F821
 @login_required
-@active_required
 @validate_request("name")
 def create():
     req = request.json
@@ -62,10 +65,40 @@ def create():
         req["name"] = dataset_name
         req["tenant_id"] = current_user.id
         req["created_by"] = current_user.id
+        if not req.get("parser_id"):
+            req["parser_id"] = "naive"
         e, t = TenantService.get_by_id(current_user.id)
         if not e:
             return get_data_error_result(message="Tenant not found.")
-        req["embd_id"] = t.embd_id
+
+        req["parser_config"] = {
+            "layout_recognize": "DeepDOC",
+            "chunk_token_num": 512,
+            "delimiter": "\n",
+            "auto_keywords": 0,
+            "auto_questions": 0,
+            "html4excel": False,
+            "topn_tags": 3,
+            "raptor": {
+                "use_raptor": True,
+                "prompt": "Please summarize the following paragraphs. Be careful with the numbers, do not make things up. Paragraphs as following:\n      {cluster_content}\nThe above is the content you need to summarize.",
+                "max_token": 256,
+                "threshold": 0.1,
+                "max_cluster": 64,
+                "random_seed": 0
+            },
+            "graphrag": {
+                "use_graphrag": True,
+                "entity_types": [
+                    "organization",
+                    "person",
+                    "geo",
+                    "event",
+                    "category"
+                ],
+                "method": "light"
+            }
+        }
         if not KnowledgebaseService.save(**req):
             return get_data_error_result()
         return get_json_result(data={"kb_id": req["id"]})
@@ -156,6 +189,9 @@ def detail():
             return get_data_error_result(
                 message="Can't find this knowledgebase!")
         kb["size"] = DocumentService.get_total_size_by_kb_id(kb_id=kb["id"],keywords="", run_status=[], types=[])
+        for key in ["graphrag_task_finish_at", "raptor_task_finish_at", "mindmap_task_finish_at"]:
+            if finish_at := kb.get(key):
+                kb[key] = finish_at.strftime("%Y-%m-%d %H:%M:%S")
         return get_json_result(data=kb)
     except Exception as e:
         return server_error_response(e)
@@ -251,7 +287,7 @@ def list_tags(kb_id):
     tenants = UserTenantService.get_tenants_by_user_id(current_user.id)
     tags = []
     for tenant in tenants:
-        tags += settings.retrievaler.all_tags(tenant["tenant_id"], [kb_id])
+        tags += settings.retriever.all_tags(tenant["tenant_id"], [kb_id])
     return get_json_result(data=tags)
 
 
@@ -270,7 +306,7 @@ def list_tags_from_kbs():
     tenants = UserTenantService.get_tenants_by_user_id(current_user.id)
     tags = []
     for tenant in tenants:
-        tags += settings.retrievaler.all_tags(tenant["tenant_id"], kb_ids)
+        tags += settings.retriever.all_tags(tenant["tenant_id"], kb_ids)
     return get_json_result(data=tags)
 
 
@@ -331,7 +367,7 @@ def knowledge_graph(kb_id):
     obj = {"graph": {}, "mind_map": {}}
     if not settings.docStoreConn.indexExist(search.index_name(kb.tenant_id), kb_id):
         return get_json_result(data=obj)
-    sres = settings.retrievaler.search(req, search.index_name(kb.tenant_id), [kb_id])
+    sres = settings.retriever.search(req, search.index_name(kb.tenant_id), [kb_id])
     if not len(sres.ids):
         return get_json_result(data=obj)
 
@@ -396,3 +432,359 @@ def get_basic_info():
     basic_info = DocumentService.knowledgebase_basic_info(kb_id)
 
     return get_json_result(data=basic_info)
+
+
+@manager.route("/list_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+def list_pipeline_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    keywords = request.args.get("keywords", "")
+
+    page_number = int(request.args.get("page", 0))
+    items_per_page = int(request.args.get("page_size", 0))
+    orderby = request.args.get("orderby", "create_time")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status status conditions: {', '.join(invalid_status)}")
+
+    types = req.get("types", [])
+    if types:
+        invalid_types = {t for t in types if t not in VALID_FILE_TYPES}
+        if invalid_types:
+            return get_data_error_result(message=f"Invalid filter conditions: {', '.join(invalid_types)} type{'s' if len(invalid_types) > 1 else ''}")
+
+    suffix = req.get("suffix", [])
+
+    try:
+        logs, tol = PipelineOperationLogService.get_file_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, keywords, operation_status, types, suffix, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/list_pipeline_dataset_logs", methods=["POST"])  # noqa: F821
+@login_required
+def list_pipeline_dataset_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    page_number = int(request.args.get("page", 0))
+    items_per_page = int(request.args.get("page_size", 0))
+    orderby = request.args.get("orderby", "create_time")
+    if request.args.get("desc", "true").lower() == "false":
+        desc = False
+    else:
+        desc = True
+    create_date_from = request.args.get("create_date_from", "")
+    create_date_to = request.args.get("create_date_to", "")
+    if create_date_to > create_date_from:
+        return get_data_error_result(message="Create data filter is abnormal.")
+
+    req = request.get_json()
+
+    operation_status = req.get("operation_status", [])
+    if operation_status:
+        invalid_status = {s for s in operation_status if s not in VALID_TASK_STATUS}
+        if invalid_status:
+            return get_data_error_result(message=f"Invalid filter operation_status status conditions: {', '.join(invalid_status)}")
+
+    try:
+        logs, tol = PipelineOperationLogService.get_dataset_logs_by_kb_id(kb_id, page_number, items_per_page, orderby, desc, operation_status, create_date_from, create_date_to)
+        return get_json_result(data={"total": tol, "logs": logs})
+    except Exception as e:
+        return server_error_response(e)
+
+
+@manager.route("/delete_pipeline_logs", methods=["POST"])  # noqa: F821
+@login_required
+def delete_pipeline_logs():
+    kb_id = request.args.get("kb_id")
+    if not kb_id:
+        return get_json_result(data=False, message='Lack of "KB ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    req = request.get_json()
+    log_ids = req.get("log_ids", [])
+
+    PipelineOperationLogService.delete_by_ids(log_ids)
+
+    return get_json_result(data=True)
+
+
+@manager.route("/pipeline_log_detail", methods=["GET"])  # noqa: F821
+@login_required
+def pipeline_log_detail():
+    log_id = request.args.get("log_id")
+    if not log_id:
+        return get_json_result(data=False, message='Lack of "Pipeline log ID"', code=settings.RetCode.ARGUMENT_ERROR)
+
+    ok, log = PipelineOperationLogService.get_by_id(log_id)
+    if not ok:
+        return get_data_error_result(message="Invalid pipeline log ID")
+
+    return get_json_result(data=log.to_dict())
+
+
+@manager.route("/run_graphrag", methods=["POST"])  # noqa: F821
+@login_required
+def run_graphrag():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid GraphRAG task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Graph Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="graphrag", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"graphrag_task_id": task_id}):
+        logging.warning(f"Cannot save graphrag_task_id for kb {kb_id}")
+
+    return get_json_result(data={"graphrag_task_id": task_id})
+
+
+@manager.route("/trace_graphrag", methods=["GET"])  # noqa: F821
+@login_required
+def trace_graphrag():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.graphrag_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="GraphRAG Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/run_raptor", methods=["POST"])  # noqa: F821
+@login_required
+def run_raptor():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid RAPTOR task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A RAPTOR Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="raptor", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"raptor_task_id": task_id}):
+        logging.warning(f"Cannot save raptor_task_id for kb {kb_id}")
+
+    return get_json_result(data={"raptor_task_id": task_id})
+
+
+@manager.route("/trace_raptor", methods=["GET"])  # noqa: F821
+@login_required
+def trace_raptor():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.raptor_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="RAPTOR Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/run_mindmap", methods=["POST"])  # noqa: F821
+@login_required
+def run_mindmap():
+    req = request.json
+
+    kb_id = req.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if task_id:
+        ok, task = TaskService.get_by_id(task_id)
+        if not ok:
+            logging.warning(f"A valid Mindmap task id is expected for kb {kb_id}")
+
+        if task and task.progress not in [-1, 1]:
+            return get_error_data_result(message=f"Task {task_id} in progress with status {task.progress}. A Mindmap Task is already running.")
+
+    documents, _ = DocumentService.get_by_kb_id(
+        kb_id=kb_id,
+        page_number=0,
+        items_per_page=0,
+        orderby="create_time",
+        desc=False,
+        keywords="",
+        run_status=[],
+        types=[],
+        suffix=[],
+    )
+    if not documents:
+        return get_error_data_result(message=f"No documents in Knowledgebase {kb_id}")
+
+    sample_document = documents[0]
+    document_ids = [document["id"] for document in documents]
+
+    task_id = queue_raptor_o_graphrag_tasks(sample_doc_id=sample_document, ty="mindmap", priority=0, fake_doc_id=GRAPH_RAPTOR_FAKE_DOC_ID, doc_ids=list(document_ids))
+
+    if not KnowledgebaseService.update_by_id(kb.id, {"mindmap_task_id": task_id}):
+        logging.warning(f"Cannot save mindmap_task_id for kb {kb_id}")
+
+    return get_json_result(data={"mindmap_task_id": task_id})
+
+
+@manager.route("/trace_mindmap", methods=["GET"])  # noqa: F821
+@login_required
+def trace_mindmap():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_error_data_result(message="Invalid Knowledgebase ID")
+
+    task_id = kb.mindmap_task_id
+    if not task_id:
+        return get_json_result(data={})
+
+    ok, task = TaskService.get_by_id(task_id)
+    if not ok:
+        return get_error_data_result(message="Mindmap Task Not Found or Error Occurred")
+
+    return get_json_result(data=task.to_dict())
+
+
+@manager.route("/unbind_task", methods=["DELETE"])  # noqa: F821
+@login_required
+def delete_kb_task():
+    kb_id = request.args.get("kb_id", "")
+    if not kb_id:
+        return get_error_data_result(message='Lack of "KB ID"')
+    ok, kb = KnowledgebaseService.get_by_id(kb_id)
+    if not ok:
+        return get_json_result(data=True)
+
+    pipeline_task_type = request.args.get("pipeline_task_type", "")
+    if not pipeline_task_type or pipeline_task_type not in [PipelineTaskType.GRAPH_RAG, PipelineTaskType.RAPTOR, PipelineTaskType.MINDMAP]:
+        return get_error_data_result(message="Invalid task type")
+
+    match pipeline_task_type:
+        case PipelineTaskType.GRAPH_RAG:
+            settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), kb_id)
+            kb_task_id_field = "graphrag_task_id"
+            task_id = kb.graphrag_task_id
+            kb_task_finish_at = "graphrag_task_finish_at"
+        case PipelineTaskType.RAPTOR:
+            kb_task_id_field = "raptor_task_id"
+            task_id = kb.raptor_task_id
+            kb_task_finish_at = "raptor_task_finish_at"
+        case PipelineTaskType.MINDMAP:
+            kb_task_id_field = "mindmap_task_id"
+            task_id = kb.mindmap_task_id
+            kb_task_finish_at = "mindmap_task_finish_at"
+        case _:
+            return get_error_data_result(message="Internal Error: Invalid task type")
+
+    def cancel_task(task_id):
+        REDIS_CONN.set(f"{task_id}-cancel", "x")
+    cancel_task(task_id)
+
+    ok = KnowledgebaseService.update_by_id(kb_id, {kb_task_id_field: "", kb_task_finish_at: None})
+    if not ok:
+        return server_error_response(f"Internal error: cannot delete task {pipeline_task_type}")
+
+    return get_json_result(data=True)

api/apps/llm_app.py

@@ -194,6 +194,9 @@ def add_llm():
     elif factory == "Azure-OpenAI":
         api_key = apikey_json(["api_key", "api_version"])
 
+    elif factory == "OpenRouter":
+        api_key = apikey_json(["api_key", "provider_order"])
+
     llm = {
         "tenant_id": current_user.id,
         "llm_factory": factory,

api/apps/plugin_app.py

@@ -1,8 +1,26 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
 from flask import Response
 from flask_login import login_required
 from api.utils.api_utils import get_json_result
 from plugin import GlobalPluginManager
 
+
 @manager.route('/llm_tools', methods=['GET'])  # noqa: F821
 @login_required
 def llm_tools() -> Response:

api/apps/sdk/agent.py

@@ -25,6 +25,7 @@ from api.utils.api_utils import get_data_error_result, get_error_data_result, ge
 from api.utils.api_utils import get_result
 from flask import request
 
+
 @manager.route('/agents', methods=['GET'])  # noqa: F821
 @token_required
 def list_agents(tenant_id):
@@ -41,7 +42,7 @@ def list_agents(tenant_id):
         desc = False
     else:
         desc = True
-    canvas = UserCanvasService.get_list(tenant_id,page_number,items_per_page,orderby,desc,id,title)
+    canvas = UserCanvasService.get_list(tenant_id, page_number, items_per_page, orderby, desc, id, title)
     return get_result(data=canvas)
 
 
@@ -93,7 +94,7 @@ def update_agent(tenant_id: str, agent_id: str):
             req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
 
         req["dsl"] = json.loads(req["dsl"])
-    
+
     if req.get("title") is not None:
         req["title"] = req["title"].strip()
 

api/apps/sdk/dataset.py

@@ -215,7 +215,8 @@ def delete(tenant_id):
                     continue
                 kb_id_instance_pairs.append((kb_id, kb))
             if len(error_kb_ids) > 0:
-                return get_error_permission_result(message=f"""User '{tenant_id}' lacks permission for datasets: '{", ".join(error_kb_ids)}'""")
+                return get_error_permission_result(
+                    message=f"""User '{tenant_id}' lacks permission for datasets: '{", ".join(error_kb_ids)}'""")
 
         errors = []
         success_count = 0
@@ -232,7 +233,8 @@ def delete(tenant_id):
                     ]
                 )
                 File2DocumentService.delete_by_document_id(doc.id)
-            FileService.filter_delete([File.source_type == FileSource.KNOWLEDGEBASE, File.type == "folder", File.name == kb.name])
+            FileService.filter_delete(
+                [File.source_type == FileSource.KNOWLEDGEBASE, File.type == "folder", File.name == kb.name])
             if not KnowledgebaseService.delete_by_id(kb_id):
                 errors.append(f"Delete dataset error for {kb_id}")
                 continue
@@ -329,7 +331,8 @@ def update(tenant_id, dataset_id):
     try:
         kb = KnowledgebaseService.get_or_none(id=dataset_id, tenant_id=tenant_id)
         if kb is None:
-            return get_error_permission_result(message=f"User '{tenant_id}' lacks permission for dataset '{dataset_id}'")
+            return get_error_permission_result(
+                message=f"User '{tenant_id}' lacks permission for dataset '{dataset_id}'")
 
         if req.get("parser_config"):
             req["parser_config"] = deep_merge(kb.parser_config, req["parser_config"])
@@ -341,7 +344,8 @@ def update(tenant_id, dataset_id):
             del req["parser_config"]
 
         if "name" in req and req["name"].lower() != kb.name.lower():
-            exists = KnowledgebaseService.get_or_none(name=req["name"], tenant_id=tenant_id, status=StatusEnum.VALID.value)
+            exists = KnowledgebaseService.get_or_none(name=req["name"], tenant_id=tenant_id,
+                                                      status=StatusEnum.VALID.value)
             if exists:
                 return get_error_data_result(message=f"Dataset name '{req['name']}' already exists")
 
@@ -349,7 +353,8 @@ def update(tenant_id, dataset_id):
             if not req["embd_id"]:
                 req["embd_id"] = kb.embd_id
             if kb.chunk_num != 0 and req["embd_id"] != kb.embd_id:
-                return get_error_data_result(message=f"When chunk_num ({kb.chunk_num}) > 0, embedding_model must remain {kb.embd_id}")
+                return get_error_data_result(
+                    message=f"When chunk_num ({kb.chunk_num}) > 0, embedding_model must remain {kb.embd_id}")
             ok, err = verify_embedding_availability(req["embd_id"], tenant_id)
             if not ok:
                 return err
@@ -359,10 +364,12 @@ def update(tenant_id, dataset_id):
                 return get_error_argument_result(message="'pagerank' can only be set when doc_engine is elasticsearch")
 
             if req["pagerank"] > 0:
-                settings.docStoreConn.update({"kb_id": kb.id}, {PAGERANK_FLD: req["pagerank"]}, search.index_name(kb.tenant_id), kb.id)
+                settings.docStoreConn.update({"kb_id": kb.id}, {PAGERANK_FLD: req["pagerank"]},
+                                             search.index_name(kb.tenant_id), kb.id)
             else:
                 # Elasticsearch requires PAGERANK_FLD be non-zero!
-                settings.docStoreConn.update({"exists": PAGERANK_FLD}, {"remove": PAGERANK_FLD}, search.index_name(kb.tenant_id), kb.id)
+                settings.docStoreConn.update({"exists": PAGERANK_FLD}, {"remove": PAGERANK_FLD},
+                                             search.index_name(kb.tenant_id), kb.id)
 
         if not KnowledgebaseService.update_by_id(kb.id, req):
             return get_error_data_result(message="Update dataset error.(Database error)")
@@ -454,7 +461,7 @@ def list_datasets(tenant_id):
                 return get_error_permission_result(message=f"User '{tenant_id}' lacks permission for dataset '{name}'")
 
         tenants = TenantService.get_joined_tenants_by_user_id(tenant_id)
-        kbs = KnowledgebaseService.get_list(
+        kbs, total = KnowledgebaseService.get_list(
             [m["tenant_id"] for m in tenants],
             tenant_id,
             args["page"],
@@ -468,14 +475,15 @@ def list_datasets(tenant_id):
         response_data_list = []
         for kb in kbs:
             response_data_list.append(remap_dictionary_keys(kb))
-        return get_result(data=response_data_list)
+        return get_result(data=response_data_list, total=total)
     except OperationalError as e:
         logging.exception(e)
         return get_error_data_result(message="Database operation failed")
 
+
 @manager.route('/datasets/<dataset_id>/knowledge_graph', methods=['GET'])  # noqa: F821
 @token_required
-def knowledge_graph(tenant_id,dataset_id):
+def knowledge_graph(tenant_id, dataset_id):
     if not KnowledgebaseService.accessible(dataset_id, tenant_id):
         return get_result(
             data=False,
@@ -491,7 +499,7 @@ def knowledge_graph(tenant_id,dataset_id):
     obj = {"graph": {}, "mind_map": {}}
     if not settings.docStoreConn.indexExist(search.index_name(kb.tenant_id), dataset_id):
         return get_result(data=obj)
-    sres = settings.retrievaler.search(req, search.index_name(kb.tenant_id), [dataset_id])
+    sres = settings.retriever.search(req, search.index_name(kb.tenant_id), [dataset_id])
     if not len(sres.ids):
         return get_result(data=obj)
 
@@ -507,14 +515,16 @@ def knowledge_graph(tenant_id,dataset_id):
     if "nodes" in obj["graph"]:
         obj["graph"]["nodes"] = sorted(obj["graph"]["nodes"], key=lambda x: x.get("pagerank", 0), reverse=True)[:256]
         if "edges" in obj["graph"]:
-            node_id_set = { o["id"] for o in obj["graph"]["nodes"] }
+            node_id_set = {o["id"] for o in obj["graph"]["nodes"]}
-            filtered_edges = [o for o in obj["graph"]["edges"] if o["source"] != o["target"] and o["source"] in node_id_set and o["target"] in node_id_set]
+            filtered_edges = [o for o in obj["graph"]["edges"] if
+                              o["source"] != o["target"] and o["source"] in node_id_set and o["target"] in node_id_set]
             obj["graph"]["edges"] = sorted(filtered_edges, key=lambda x: x.get("weight", 0), reverse=True)[:128]
     return get_result(data=obj)
 
+
 @manager.route('/datasets/<dataset_id>/knowledge_graph', methods=['DELETE'])  # noqa: F821
 @token_required
-def delete_knowledge_graph(tenant_id,dataset_id):
+def delete_knowledge_graph(tenant_id, dataset_id):
     if not KnowledgebaseService.accessible(dataset_id, tenant_id):
         return get_result(
             data=False,
@@ -522,6 +532,7 @@ def delete_knowledge_graph(tenant_id,dataset_id):
             code=settings.RetCode.AUTHENTICATION_ERROR
         )
     _, kb = KnowledgebaseService.get_by_id(dataset_id)
-    settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]}, search.index_name(kb.tenant_id), dataset_id)
+    settings.docStoreConn.delete({"knowledge_graph_kwd": ["graph", "subgraph", "entity", "relation"]},
+                                 search.index_name(kb.tenant_id), dataset_id)
 
     return get_result(data=True)

api/apps/sdk/dify_retrieval.py

@@ -1,4 +1,4 @@
-    #
+#
 #  Copyright 2024 The InfiniFlow Authors. All Rights Reserved.
 #
 #  Licensed under the Apache License, Version 2.0 (the "License");
@@ -31,6 +31,89 @@ from api.db.services.dialog_service import meta_filter, convert_conditions
 @apikey_required
 @validate_request("knowledge_id", "query")
 def retrieval(tenant_id):
+    """
+    Dify-compatible retrieval API
+    ---
+    tags:
+      - SDK
+    security:
+      - ApiKeyAuth: []
+    parameters:
+      - in: body
+        name: body
+        required: true
+        schema:
+          type: object
+          required:
+            - knowledge_id
+            - query
+          properties:
+            knowledge_id:
+              type: string
+              description: Knowledge base ID
+            query:
+              type: string
+              description: Query text
+            use_kg:
+              type: boolean
+              description: Whether to use knowledge graph
+              default: false
+            retrieval_setting:
+              type: object
+              description: Retrieval configuration
+              properties:
+                score_threshold:
+                  type: number
+                  description: Similarity threshold
+                  default: 0.0
+                top_k:
+                  type: integer
+                  description: Number of results to return
+                  default: 1024
+            metadata_condition:
+              type: object
+              description: Metadata filter condition
+              properties:
+                conditions:
+                  type: array
+                  items:
+                    type: object
+                    properties:
+                      name:
+                        type: string
+                        description: Field name
+                      comparison_operator:
+                        type: string
+                        description: Comparison operator
+                      value:
+                        type: string
+                        description: Field value
+    responses:
+      200:
+        description: Retrieval succeeded
+        schema:
+          type: object
+          properties:
+            records:
+              type: array
+              items:
+                type: object
+                properties:
+                  content:
+                    type: string
+                    description: Content text
+                  score:
+                    type: number
+                    description: Similarity score
+                  title:
+                    type: string
+                    description: Document title
+                  metadata:
+                    type: object
+                    description: Metadata info
+      404:
+        description: Knowledge base or document not found
+    """
     req = request.json
     question = req["query"]
     kb_id = req["knowledge_id"]
@@ -38,9 +121,9 @@ def retrieval(tenant_id):
     retrieval_setting = req.get("retrieval_setting", {})
     similarity_threshold = float(retrieval_setting.get("score_threshold", 0.0))
     top = int(retrieval_setting.get("top_k", 1024))
-    metadata_condition = req.get("metadata_condition",{})
+    metadata_condition = req.get("metadata_condition", {})
     metas = DocumentService.get_meta_by_kbs([kb_id])
- 
+
     doc_ids = []
     try:
 
@@ -50,12 +133,12 @@ def retrieval(tenant_id):
 
         embd_mdl = LLMBundle(kb.tenant_id, LLMType.EMBEDDING.value, llm_name=kb.embd_id)
         print(metadata_condition)
-        print("after",convert_conditions(metadata_condition))
+        # print("after", convert_conditions(metadata_condition))
         doc_ids.extend(meta_filter(metas, convert_conditions(metadata_condition)))
-        print("doc_ids",doc_ids)
+        # print("doc_ids", doc_ids)
         if not doc_ids and metadata_condition is not None:
             doc_ids = ['-999']
-        ranks = settings.retrievaler.retrieval(
+        ranks = settings.retriever.retrieval(
             question,
             embd_mdl,
             kb.tenant_id,
@@ -70,17 +153,17 @@ def retrieval(tenant_id):
         )
 
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question,
+            ck = settings.kg_retriever.retrieval(question,
-                                                   [tenant_id],
+                                                 [tenant_id],
-                                                   [kb_id],
+                                                 [kb_id],
-                                                   embd_mdl,
+                                                 embd_mdl,
-                                                   LLMBundle(kb.tenant_id, LLMType.CHAT))
+                                                 LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
 
         records = []
         for c in ranks["chunks"]:
-            e, doc = DocumentService.get_by_id( c["doc_id"])
+            e, doc = DocumentService.get_by_id(c["doc_id"])
             c.pop("vector", None)
             meta = getattr(doc, 'meta_fields', {})
             meta["doc_id"] = c["doc_id"]
@@ -100,5 +183,3 @@ def retrieval(tenant_id):
             )
         logging.exception(e)
         return build_error_result(message=str(e), code=settings.RetCode.SERVER_ERROR)
-
-

api/apps/sdk/doc.py

@@ -458,7 +458,7 @@ def list_docs(dataset_id, tenant_id):
         required: false
         default: true
         description: Order in descending.
-    - in: query
+      - in: query
         name: create_time_from
         type: integer
         required: false
@@ -470,6 +470,20 @@ def list_docs(dataset_id, tenant_id):
         required: false
         default: 0
         description: Unix timestamp for filtering documents created before this time. 0 means no filter.
+      - in: query
+        name: suffix
+        type: array
+        items:
+          type: string
+        required: false
+        description: Filter by file suffix (e.g., ["pdf", "txt", "docx"]).
+      - in: query
+        name: run
+        type: array
+        items:
+          type: string
+        required: false
+        description: Filter by document run status. Supports both numeric ("0", "1", "2", "3", "4") and text formats ("UNSTART", "RUNNING", "CANCEL", "DONE", "FAIL").
       - in: header
         name: Authorization
         type: string
@@ -512,63 +526,62 @@ def list_docs(dataset_id, tenant_id):
                     description: Processing status.
     """
     if not KnowledgebaseService.accessible(kb_id=dataset_id, user_id=tenant_id):
-        return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
+      return get_error_data_result(message=f"You don't own the dataset {dataset_id}. ")
-    id = request.args.get("id")
-    name = request.args.get("name")
 
-    if id and not DocumentService.query(id=id, kb_id=dataset_id):
+    q = request.args
-        return get_error_data_result(message=f"You don't own the document {id}.")
+    document_id = q.get("id")  
+    name        = q.get("name")
+
+    if document_id and not DocumentService.query(id=document_id, kb_id=dataset_id):
+        return get_error_data_result(message=f"You don't own the document {document_id}.")
     if name and not DocumentService.query(name=name, kb_id=dataset_id):
         return get_error_data_result(message=f"You don't own the document {name}.")
 
-    page = int(request.args.get("page", 1))
+    page        = int(q.get("page", 1))
-    keywords = request.args.get("keywords", "")
+    page_size   = int(q.get("page_size", 30))  
-    page_size = int(request.args.get("page_size", 30))
+    orderby     = q.get("orderby", "create_time")
-    orderby = request.args.get("orderby", "create_time")
+    desc        = str(q.get("desc", "true")).strip().lower() != "false"
-    if request.args.get("desc") == "False":
+    keywords    = q.get("keywords", "")
-        desc = False
-    else:
-        desc = True
-    docs, tol = DocumentService.get_list(dataset_id, page, page_size, orderby, desc, keywords, id, name)
 
-    create_time_from = int(request.args.get("create_time_from", 0))
+    # filters - align with OpenAPI parameter names
-    create_time_to = int(request.args.get("create_time_to", 0))
+    suffix               = q.getlist("suffix") 
+    run_status           = q.getlist("run")   
+    create_time_from     = int(q.get("create_time_from", 0))  
+    create_time_to       = int(q.get("create_time_to", 0))    
 
+    # map run status (accept text or numeric) - align with API parameter
+    run_status_text_to_numeric = {"UNSTART": "0", "RUNNING": "1", "CANCEL": "2", "DONE": "3", "FAIL": "4"}
+    run_status_converted = [run_status_text_to_numeric.get(v, v) for v in run_status]
+
+    docs, total = DocumentService.get_list(
+        dataset_id, page, page_size, orderby, desc, keywords, document_id, name, suffix, run_status_converted
+    )
+
+    # time range filter (0 means no bound)
     if create_time_from or create_time_to:
-        filtered_docs = []
+        docs = [
-        for doc in docs:
+            d for d in docs
-            doc_create_time = doc.get("create_time", 0)
+            if (create_time_from == 0 or d.get("create_time", 0) >= create_time_from)
-            if (create_time_from == 0 or doc_create_time >= create_time_from) and (create_time_to == 0 or doc_create_time <= create_time_to):
+            and (create_time_to == 0 or d.get("create_time", 0) <= create_time_to)
-                filtered_docs.append(doc)
+        ]
-        docs = filtered_docs
 
-    # rename key's name
+    # rename keys + map run status back to text for output
-    renamed_doc_list = []
     key_mapping = {
         "chunk_num": "chunk_count",
-        "kb_id": "dataset_id",
+        "kb_id": "dataset_id", 
         "token_num": "token_count",
         "parser_id": "chunk_method",
     }
-    run_mapping = {
+    run_status_numeric_to_text = {"0": "UNSTART", "1": "RUNNING", "2": "CANCEL", "3": "DONE", "4": "FAIL"}
-        "0": "UNSTART",
-        "1": "RUNNING",
-        "2": "CANCEL",
-        "3": "DONE",
-        "4": "FAIL",
-    }
-    for doc in docs:
-        renamed_doc = {}
-        for key, value in doc.items():
-            if key == "run":
-                renamed_doc["run"] = run_mapping.get(str(value))
-            new_key = key_mapping.get(key, key)
-            renamed_doc[new_key] = value
-            if key == "run":
-                renamed_doc["run"] = run_mapping.get(value)
-        renamed_doc_list.append(renamed_doc)
-    return get_result(data={"total": tol, "docs": renamed_doc_list})
 
+    output_docs = []
+    for d in docs:
+        renamed_doc = {key_mapping.get(k, k): v for k, v in d.items()}
+        if "run" in d:
+            renamed_doc["run"] = run_status_numeric_to_text.get(str(d["run"]), d["run"])
+        output_docs.append(renamed_doc)
+
+    return get_result(data={"total": total, "docs": output_docs})
 
 @manager.route("/datasets/<dataset_id>/documents", methods=["DELETE"])  # noqa: F821
 @token_required
@@ -982,7 +995,7 @@ def list_chunks(tenant_id, dataset_id, document_id):
         _ = Chunk(**final_chunk)
 
     elif settings.docStoreConn.indexExist(search.index_name(tenant_id), dataset_id):
-        sres = settings.retrievaler.search(query, search.index_name(tenant_id), [dataset_id], emb_mdl=None, highlight=True)
+        sres = settings.retriever.search(query, search.index_name(tenant_id), [dataset_id], emb_mdl=None, highlight=True)
         res["total"] = sres.total
         for id in sres.ids:
             d = {
@@ -1446,7 +1459,7 @@ def retrieval_test(tenant_id):
             chat_mdl = LLMBundle(kb.tenant_id, LLMType.CHAT)
             question += keyword_extraction(chat_mdl, question)
 
-        ranks = settings.retrievaler.retrieval(
+        ranks = settings.retriever.retrieval(
             question,
             embd_mdl,
             tenant_ids,
@@ -1462,7 +1475,7 @@ def retrieval_test(tenant_id):
             rank_feature=label_question(question, kbs),
         )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question, [k.tenant_id for k in kbs], kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
+            ck = settings.kg_retriever.retrieval(question, [k.tenant_id for k in kbs], kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
 

api/apps/sdk/files.py

@@ -1,3 +1,20 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
 import pathlib
 import re
 
@@ -17,7 +34,8 @@ from api.utils.api_utils import get_json_result
 from api.utils.file_utils import filename_type
 from rag.utils.storage_factory import STORAGE_IMPL
 
-@manager.route('/file/upload', methods=['POST']) # noqa: F821
+
+@manager.route('/file/upload', methods=['POST'])  # noqa: F821
 @token_required
 def upload(tenant_id):
     """
@@ -44,22 +62,22 @@ def upload(tenant_id):
           type: object
           properties:
             data:
-            type: array
+              type: array
-            items:
+              items:
-              type: object
+                type: object
-              properties:
+                properties:
-                id:
+                  id:
-                  type: string
+                    type: string
-                  description: File ID
+                    description: File ID
-                name:
+                  name:
-                  type: string
+                    type: string
-                  description: File name
+                    description: File name
-                size:
+                  size:
-                  type: integer
+                    type: integer
-                  description: File size in bytes
+                    description: File size in bytes
-                type:
+                  type:
-                  type: string
+                    type: string
-                  description: File type (e.g., document, folder)
+                    description: File type (e.g., document, folder)
     """
     pf_id = request.form.get("parent_id")
 
@@ -97,12 +115,14 @@ def upload(tenant_id):
                 e, file = FileService.get_by_id(file_id_list[len_id_list - 1])
                 if not e:
                     return get_json_result(data=False, message="Folder not found!", code=404)
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 1], file_obj_names, len_id_list)
+                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 1], file_obj_names,
+                                                        len_id_list)
             else:
                 e, file = FileService.get_by_id(file_id_list[len_id_list - 2])
                 if not e:
                     return get_json_result(data=False, message="Folder not found!", code=404)
-                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 2], file_obj_names, len_id_list)
+                last_folder = FileService.create_folder(file, file_id_list[len_id_list - 2], file_obj_names,
+                                                        len_id_list)
 
             filetype = filename_type(file_obj_names[file_len - 1])
             location = file_obj_names[file_len - 1]
@@ -129,7 +149,7 @@ def upload(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/create', methods=['POST']) # noqa: F821
+@manager.route('/file/create', methods=['POST'])  # noqa: F821
 @token_required
 def create(tenant_id):
     """
@@ -207,7 +227,7 @@ def create(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/list', methods=['GET']) # noqa: F821
+@manager.route('/file/list', methods=['GET'])  # noqa: F821
 @token_required
 def list_files(tenant_id):
     """
@@ -299,7 +319,7 @@ def list_files(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/root_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/root_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_root_folder(tenant_id):
     """
@@ -335,7 +355,7 @@ def get_root_folder(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/parent_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/parent_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_parent_folder():
     """
@@ -380,7 +400,7 @@ def get_parent_folder():
         return server_error_response(e)
 
 
-@manager.route('/file/all_parent_folder', methods=['GET']) # noqa: F821
+@manager.route('/file/all_parent_folder', methods=['GET'])  # noqa: F821
 @token_required
 def get_all_parent_folders(tenant_id):
     """
@@ -428,7 +448,7 @@ def get_all_parent_folders(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/rm', methods=['POST']) # noqa: F821
+@manager.route('/file/rm', methods=['POST'])  # noqa: F821
 @token_required
 def rm(tenant_id):
     """
@@ -502,7 +522,7 @@ def rm(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/rename', methods=['POST']) # noqa: F821
+@manager.route('/file/rename', methods=['POST'])  # noqa: F821
 @token_required
 def rename(tenant_id):
     """
@@ -542,7 +562,8 @@ def rename(tenant_id):
         if not e:
             return get_json_result(message="File not found!", code=404)
 
-        if file.type != FileType.FOLDER.value and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(file.name.lower()).suffix:
+        if file.type != FileType.FOLDER.value and pathlib.Path(req["name"].lower()).suffix != pathlib.Path(
+                file.name.lower()).suffix:
             return get_json_result(data=False, message="The extension of file can't be changed", code=400)
 
         for existing_file in FileService.query(name=req["name"], pf_id=file.parent_id):
@@ -562,9 +583,9 @@ def rename(tenant_id):
         return server_error_response(e)
 
 
-@manager.route('/file/get/<file_id>', methods=['GET']) # noqa: F821
+@manager.route('/file/get/<file_id>', methods=['GET'])  # noqa: F821
 @token_required
-def get(tenant_id,file_id):
+def get(tenant_id, file_id):
     """
     Download a file.
     ---
@@ -610,7 +631,7 @@ def get(tenant_id,file_id):
         return server_error_response(e)
 
 
-@manager.route('/file/mv', methods=['POST']) # noqa: F821
+@manager.route('/file/mv', methods=['POST'])  # noqa: F821
 @token_required
 def move(tenant_id):
     """
@@ -669,6 +690,7 @@ def move(tenant_id):
     except Exception as e:
         return server_error_response(e)
 
+
 @manager.route('/file/convert', methods=['POST'])  # noqa: F821
 @token_required
 def convert(tenant_id):
@@ -735,4 +757,4 @@ def convert(tenant_id):
                     file2documents.append(file2document.to_json())
         return get_json_result(data=file2documents)
     except Exception as e:
-        return server_error_response(e)
+        return server_error_response(e)

api/apps/sdk/session.py

@@ -36,7 +36,8 @@ from api.db.services.llm_service import LLMBundle
 from api.db.services.search_service import SearchService
 from api.db.services.user_service import UserTenantService
 from api.utils import get_uuid
-from api.utils.api_utils import check_duplicate_ids, get_data_openai, get_error_data_result, get_json_result, get_result, server_error_response, token_required, validate_request
+from api.utils.api_utils import check_duplicate_ids, get_data_openai, get_error_data_result, get_json_result, \
+    get_result, server_error_response, token_required, validate_request
 from rag.app.tag import label_question
 from rag.prompts.template import load_prompt
 from rag.prompts.generator import cross_languages, gen_meta_filter, keyword_extraction, chunks_format
@@ -88,7 +89,8 @@ def create_agent_session(tenant_id, agent_id):
     canvas.reset()
 
     cvs.dsl = json.loads(str(canvas))
-    conv = {"id": session_id, "dialog_id": cvs.id, "user_id": user_id, "message": [{"role": "assistant", "content": canvas.get_prologue()}], "source": "agent", "dsl": cvs.dsl}
+    conv = {"id": session_id, "dialog_id": cvs.id, "user_id": user_id,
+            "message": [{"role": "assistant", "content": canvas.get_prologue()}], "source": "agent", "dsl": cvs.dsl}
     API4ConversationService.save(**conv)
     conv["agent_id"] = conv.pop("dialog_id")
     return get_result(data=conv)
@@ -279,7 +281,7 @@ def chat_completion_openai_like(tenant_id, chat_id):
                     reasoning_match = re.search(r"<think>(.*?)</think>", answer, flags=re.DOTALL)
                     if reasoning_match:
                         reasoning_part = reasoning_match.group(1)
-                        content_part = answer[reasoning_match.end() :]
+                        content_part = answer[reasoning_match.end():]
                     else:
                         reasoning_part = ""
                         content_part = answer
@@ -324,7 +326,8 @@ def chat_completion_openai_like(tenant_id, chat_id):
             response["choices"][0]["delta"]["content"] = None
             response["choices"][0]["delta"]["reasoning_content"] = None
             response["choices"][0]["finish_reason"] = "stop"
-            response["usage"] = {"prompt_tokens": len(prompt), "completion_tokens": token_used, "total_tokens": len(prompt) + token_used}
+            response["usage"] = {"prompt_tokens": len(prompt), "completion_tokens": token_used,
+                                 "total_tokens": len(prompt) + token_used}
             if need_reference:
                 response["choices"][0]["delta"]["reference"] = chunks_format(last_ans.get("reference", []))
                 response["choices"][0]["delta"]["final_content"] = last_ans.get("answer", "")
@@ -559,7 +562,8 @@ def list_agent_session(tenant_id, agent_id):
         desc = True
     # dsl defaults to True in all cases except for False and false
     include_dsl = request.args.get("dsl") != "False" and request.args.get("dsl") != "false"
-    total, convs = API4ConversationService.get_list(agent_id, tenant_id, page_number, items_per_page, orderby, desc, id, user_id, include_dsl)
+    total, convs = API4ConversationService.get_list(agent_id, tenant_id, page_number, items_per_page, orderby, desc, id,
+                                                    user_id, include_dsl)
     if not convs:
         return get_result(data=[])
     for conv in convs:
@@ -581,7 +585,8 @@ def list_agent_session(tenant_id, agent_id):
                 if message_num != 0 and messages[message_num]["role"] != "user":
                     chunk_list = []
                     # Add boundary and type checks to prevent KeyError
-                    if chunk_num < len(conv["reference"]) and conv["reference"][chunk_num] is not None and isinstance(conv["reference"][chunk_num], dict) and "chunks" in conv["reference"][chunk_num]:
+                    if chunk_num < len(conv["reference"]) and conv["reference"][chunk_num] is not None and isinstance(
+                            conv["reference"][chunk_num], dict) and "chunks" in conv["reference"][chunk_num]:
                         chunks = conv["reference"][chunk_num]["chunks"]
                         for chunk in chunks:
                             # Ensure chunk is a dictionary before calling get method
@@ -639,13 +644,16 @@ def delete(tenant_id, chat_id):
 
     if errors:
         if success_count > 0:
-            return get_result(data={"success_count": success_count, "errors": errors}, message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
+            return get_result(data={"success_count": success_count, "errors": errors},
+                              message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
         else:
             return get_error_data_result(message="; ".join(errors))
 
     if duplicate_messages:
         if success_count > 0:
-            return get_result(message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors", data={"success_count": success_count, "errors": duplicate_messages})
+            return get_result(
+                message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors",
+                data={"success_count": success_count, "errors": duplicate_messages})
         else:
             return get_error_data_result(message=";".join(duplicate_messages))
 
@@ -691,13 +699,16 @@ def delete_agent_session(tenant_id, agent_id):
 
     if errors:
         if success_count > 0:
-            return get_result(data={"success_count": success_count, "errors": errors}, message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
+            return get_result(data={"success_count": success_count, "errors": errors},
+                              message=f"Partially deleted {success_count} sessions with {len(errors)} errors")
         else:
             return get_error_data_result(message="; ".join(errors))
 
     if duplicate_messages:
         if success_count > 0:
-            return get_result(message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors", data={"success_count": success_count, "errors": duplicate_messages})
+            return get_result(
+                message=f"Partially deleted {success_count} sessions with {len(duplicate_messages)} errors",
+                data={"success_count": success_count, "errors": duplicate_messages})
         else:
             return get_error_data_result(message=";".join(duplicate_messages))
 
@@ -730,7 +741,9 @@ def ask_about(tenant_id):
             for ans in ask(req["question"], req["kb_ids"], uid):
                 yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
         except Exception as e:
-            yield "data:" + json.dumps({"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}}, ensure_ascii=False) + "\n\n"
+            yield "data:" + json.dumps(
+                {"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
+                ensure_ascii=False) + "\n\n"
         yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
 
     resp = Response(stream(), mimetype="text/event-stream")
@@ -882,7 +895,9 @@ def begin_inputs(agent_id):
         return get_error_data_result(f"Can't find agent by ID: {agent_id}")
 
     canvas = Canvas(json.dumps(cvs.dsl), objs[0].tenant_id)
-    return get_result(data={"title": cvs.title, "avatar": cvs.avatar, "inputs": canvas.get_component_input_form("begin"), "prologue": canvas.get_prologue(), "mode": canvas.get_mode()})
+    return get_result(
+        data={"title": cvs.title, "avatar": cvs.avatar, "inputs": canvas.get_component_input_form("begin"),
+              "prologue": canvas.get_prologue(), "mode": canvas.get_mode()})
 
 
 @manager.route("/searchbots/ask", methods=["POST"])  # noqa: F821
@@ -911,7 +926,9 @@ def ask_about_embedded():
             for ans in ask(req["question"], req["kb_ids"], uid, search_config=search_config):
                 yield "data:" + json.dumps({"code": 0, "message": "", "data": ans}, ensure_ascii=False) + "\n\n"
         except Exception as e:
-            yield "data:" + json.dumps({"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}}, ensure_ascii=False) + "\n\n"
+            yield "data:" + json.dumps(
+                {"code": 500, "message": str(e), "data": {"answer": "**ERROR**: " + str(e), "reference": []}},
+                ensure_ascii=False) + "\n\n"
         yield "data:" + json.dumps({"code": 0, "message": "", "data": True}, ensure_ascii=False) + "\n\n"
 
     resp = Response(stream(), mimetype="text/event-stream")
@@ -978,7 +995,8 @@ def retrieval_test_embedded():
                     tenant_ids.append(tenant.tenant_id)
                     break
             else:
-                return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.", code=settings.RetCode.OPERATING_ERROR)
+                return get_json_result(data=False, message="Only owner of knowledgebase authorized for this operation.",
+                                       code=settings.RetCode.OPERATING_ERROR)
 
         e, kb = KnowledgebaseService.get_by_id(kb_ids[0])
         if not e:
@@ -998,11 +1016,13 @@ def retrieval_test_embedded():
             question += keyword_extraction(chat_mdl, question)
 
         labels = label_question(question, [kb])
-        ranks = settings.retrievaler.retrieval(
+        ranks = settings.retriever.retrieval(
-            question, embd_mdl, tenant_ids, kb_ids, page, size, similarity_threshold, vector_similarity_weight, top, doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"), rank_feature=labels
+            question, embd_mdl, tenant_ids, kb_ids, page, size, similarity_threshold, vector_similarity_weight, top,
+            doc_ids, rerank_mdl=rerank_mdl, highlight=req.get("highlight"), rank_feature=labels
         )
         if use_kg:
-            ck = settings.kg_retrievaler.retrieval(question, tenant_ids, kb_ids, embd_mdl, LLMBundle(kb.tenant_id, LLMType.CHAT))
+            ck = settings.kg_retriever.retrieval(question, tenant_ids, kb_ids, embd_mdl,
+                                                 LLMBundle(kb.tenant_id, LLMType.CHAT))
             if ck["content_with_weight"]:
                 ranks["chunks"].insert(0, ck)
 
@@ -1013,7 +1033,8 @@ def retrieval_test_embedded():
         return get_json_result(data=ranks)
     except Exception as e:
         if str(e).find("not_found") > 0:
-            return get_json_result(data=False, message="No chunk found! Check the chunk status please!", code=settings.RetCode.DATA_ERROR)
+            return get_json_result(data=False, message="No chunk found! Check the chunk status please!",
+                                   code=settings.RetCode.DATA_ERROR)
         return server_error_response(e)
 
 
@@ -1082,7 +1103,8 @@ def detail_share_embedded():
             if SearchService.query(tenant_id=tenant.tenant_id, id=search_id):
                 break
         else:
-            return get_json_result(data=False, message="Has no permission for this operation.", code=settings.RetCode.OPERATING_ERROR)
+            return get_json_result(data=False, message="Has no permission for this operation.",
+                                   code=settings.RetCode.OPERATING_ERROR)
 
         search = SearchService.get_detail(search_id)
         if not search:

api/apps/system_app.py

@@ -39,6 +39,7 @@ from rag.utils.redis_conn import REDIS_CONN
 from flask import jsonify
 from api.utils.health_utils import run_health_checks
 
+
 @manager.route("/version", methods=["GET"])  # noqa: F821
 @login_required
 def version():
@@ -161,7 +162,7 @@ def status():
         task_executors = REDIS_CONN.smembers("TASKEXE")
         now = datetime.now().timestamp()
         for task_executor_id in task_executors:
-            heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60*30, now)
+            heartbeats = REDIS_CONN.zrangebyscore(task_executor_id, now - 60 * 30, now)
             heartbeats = [json.loads(heartbeat) for heartbeat in heartbeats]
             task_executor_heartbeats[task_executor_id] = heartbeats
     except Exception:
@@ -177,6 +178,11 @@ def healthz():
     return jsonify(result), (200 if all_ok else 500)
 
 
+@manager.route("/ping", methods=["GET"]) # noqa: F821
+def ping():
+    return "pong", 200
+
+
 @manager.route("/new_token", methods=["POST"])  # noqa: F821
 @login_required
 def new_token():
@@ -268,7 +274,8 @@ def token_list():
         objs = [o.to_dict() for o in objs]
         for o in objs:
             if not o["beta"]:
-                o["beta"] = generate_confirmation_token(generate_confirmation_token(tenants[0].tenant_id)).replace("ragflow-", "")[:32]
+                o["beta"] = generate_confirmation_token(generate_confirmation_token(tenants[0].tenant_id)).replace(
+                    "ragflow-", "")[:32]
                 APITokenService.filter_update([APIToken.tenant_id == tenant_id, APIToken.token == o["token"]], o)
         return get_json_result(data=objs)
     except Exception as e:

api/apps/tenant_app.py

@@ -70,7 +70,8 @@ def create(tenant_id):
             return get_data_error_result(message=f"{invite_user_email} is already in the team.")
         if user_tenant_role == UserTenantRole.OWNER:
             return get_data_error_result(message=f"{invite_user_email} is the owner of the team.")
-        return get_data_error_result(message=f"{invite_user_email} is in the team, but the role: {user_tenant_role} is invalid.")
+        return get_data_error_result(
+            message=f"{invite_user_email} is in the team, but the role: {user_tenant_role} is invalid.")
 
     UserTenantService.save(
         id=get_uuid(),
@@ -132,7 +133,8 @@ def tenant_list():
 @login_required
 def agree(tenant_id):
     try:
-        UserTenantService.filter_update([UserTenant.tenant_id == tenant_id, UserTenant.user_id == current_user.id], {"role": UserTenantRole.NORMAL})
+        UserTenantService.filter_update([UserTenant.tenant_id == tenant_id, UserTenant.user_id == current_user.id],
+                                        {"role": UserTenantRole.NORMAL})
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)

api/apps/user_app.py

@@ -15,11 +15,14 @@
 #
 import json
 import logging
+import string
+import os
 import re
 import secrets
+import time
 from datetime import datetime
 
-from flask import redirect, request, session
+from flask import redirect, request, session, make_response
 from flask_login import current_user, login_required, login_user, logout_user
 from werkzeug.security import check_password_hash, generate_password_hash
 
@@ -46,6 +49,19 @@ from api.utils.api_utils import (
     validate_request,
 )
 from api.utils.crypt import decrypt
+from rag.utils.redis_conn import REDIS_CONN
+from api.apps import smtp_mail_server
+from api.utils.web_utils import (
+    send_email_html,
+    OTP_LENGTH,
+    OTP_TTL_SECONDS,
+    ATTEMPT_LIMIT,
+    ATTEMPT_LOCK_SECONDS,
+    RESEND_COOLDOWN_SECONDS,
+    otp_keys,
+    hash_code,
+    captcha_key,
+)
 
 
 @manager.route("/login", methods=["POST", "GET"])  # noqa: F821
@@ -98,7 +114,14 @@ def login():
         return get_json_result(data=False, code=settings.RetCode.SERVER_ERROR, message="Fail to crypt password")
 
     user = UserService.query_user(email, password)
-    if user:
+
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return get_json_result(
+            data=False,
+            code=settings.RetCode.FORBIDDEN,
+            message="This account has been disabled, please contact the administrator!",
+        )
+    elif user:
         response_data = user.to_json()
         user.access_token = get_uuid()
         login_user(user)
@@ -227,6 +250,9 @@ def oauth_callback(channel):
         # User exists, try to log in
         user = users[0]
         user.access_token = get_uuid()
+        if user and hasattr(user, 'is_active') and user.is_active == "0":
+            return redirect("/?error=user_inactive")
+
         login_user(user)
         user.save()
         return redirect(f"/?auth={user.get_id()}")
@@ -317,6 +343,8 @@ def github_callback():
     # User has already registered, try to log in
     user = users[0]
     user.access_token = get_uuid()
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return redirect("/?error=user_inactive")
     login_user(user)
     user.save()
     return redirect("/?auth=%s" % user.get_id())
@@ -418,6 +446,8 @@ def feishu_callback():
 
     # User has already registered, try to log in
     user = users[0]
+    if user and hasattr(user, 'is_active') and user.is_active == "0":
+        return redirect("/?error=user_inactive")
     user.access_token = get_uuid()
     login_user(user)
     user.save()
@@ -811,3 +841,172 @@ def set_tenant_info():
         return get_json_result(data=True)
     except Exception as e:
         return server_error_response(e)
+        
+
+@manager.route("/forget/captcha", methods=["GET"])  # noqa: F821
+def forget_get_captcha():
+    """
+    GET /forget/captcha?email=<email>
+    - Generate an image captcha and cache it in Redis under key captcha:{email} with TTL = OTP_TTL_SECONDS.
+    - Returns the captcha as a PNG image.
+    """
+    email = (request.args.get("email") or "")
+    if not email:
+        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email is required")
+
+    users = UserService.query(email=email)
+    if not users:
+        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+
+    # Generate captcha text
+    allowed = string.ascii_uppercase + string.digits
+    captcha_text = "".join(secrets.choice(allowed) for _ in range(OTP_LENGTH))
+    REDIS_CONN.set(captcha_key(email), captcha_text, 60) # Valid for 60 seconds
+
+    from captcha.image import ImageCaptcha
+    image = ImageCaptcha(width=300, height=120, font_sizes=[50, 60, 70])
+    img_bytes = image.generate(captcha_text).read()
+    response = make_response(img_bytes)
+    response.headers.set("Content-Type", "image/JPEG")
+    return response
+
+
+@manager.route("/forget/otp", methods=["POST"])  # noqa: F821
+def forget_send_otp():
+    """
+    POST /forget/otp
+    - Verify the image captcha stored at captcha:{email} (case-insensitive).
+    - On success, generate an email OTP (A–Z with length = OTP_LENGTH), store hash + salt (and timestamp) in Redis with TTL, reset attempts and cooldown, and send the OTP via email.
+    """
+    req = request.get_json()
+    email = req.get("email") or ""
+    captcha = (req.get("captcha") or "").strip()
+
+    if not email or not captcha:
+        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email and captcha required")
+
+    users = UserService.query(email=email)
+    if not users:
+        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+
+    stored_captcha = REDIS_CONN.get(captcha_key(email))
+    if not stored_captcha:
+        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="invalid or expired captcha")
+    if (stored_captcha or "").strip().lower() != captcha.lower():
+        return get_json_result(data=False, code=settings.RetCode.AUTHENTICATION_ERROR, message="invalid or expired captcha")
+
+    # Delete captcha to prevent reuse
+    REDIS_CONN.delete(captcha_key(email))
+
+    k_code, k_attempts, k_last, k_lock = otp_keys(email)
+    now = int(time.time())
+    last_ts = REDIS_CONN.get(k_last)
+    if last_ts:
+        try:
+            elapsed = now - int(last_ts)
+        except Exception:
+            elapsed = RESEND_COOLDOWN_SECONDS
+        remaining = RESEND_COOLDOWN_SECONDS - elapsed
+        if remaining > 0:
+            return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message=f"you still have to wait {remaining} seconds")
+
+    # Generate OTP (uppercase letters only) and store hashed
+    otp = "".join(secrets.choice(string.ascii_uppercase) for _ in range(OTP_LENGTH))
+    salt = os.urandom(16)
+    code_hash = hash_code(otp, salt)
+    REDIS_CONN.set(k_code, f"{code_hash}:{salt.hex()}", OTP_TTL_SECONDS)
+    REDIS_CONN.set(k_attempts, 0, OTP_TTL_SECONDS)
+    REDIS_CONN.set(k_last, now, OTP_TTL_SECONDS)
+    REDIS_CONN.delete(k_lock)
+
+    ttl_min = OTP_TTL_SECONDS // 60
+
+    if not smtp_mail_server:
+        logging.warning("SMTP mail server not initialized; skip sending email.")
+    else:
+        try:
+            send_email_html(
+                subject="Your Password Reset Code",
+                to_email=email,
+                template_key="reset_code",
+                code=otp,
+                ttl_min=ttl_min,
+            )
+        except Exception:
+            return get_json_result(data=False, code=settings.RetCode.SERVER_ERROR, message="failed to send email")
+        
+    return get_json_result(data=True, code=settings.RetCode.SUCCESS, message="verification passed, email sent")
+
+
+@manager.route("/forget", methods=["POST"])  # noqa: F821
+def forget():
+    """
+    POST: Verify email + OTP and reset password, then log the user in.
+    Request JSON: { email, otp, new_password, confirm_new_password }
+    """
+    req = request.get_json()
+    email = req.get("email") or ""
+    otp = (req.get("otp") or "").strip()
+    new_pwd = req.get("new_password")
+    new_pwd2 = req.get("confirm_new_password")
+
+    if not all([email, otp, new_pwd, new_pwd2]):
+        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="email, otp and passwords are required")
+
+    # For reset, passwords are provided as-is (no decrypt needed)
+    if new_pwd != new_pwd2:
+        return get_json_result(data=False, code=settings.RetCode.ARGUMENT_ERROR, message="passwords do not match")
+
+    users = UserService.query(email=email)
+    if not users:
+        return get_json_result(data=False, code=settings.RetCode.DATA_ERROR, message="invalid email")
+
+    user = users[0]
+    # Verify OTP from Redis
+    k_code, k_attempts, k_last, k_lock = otp_keys(email)
+    if REDIS_CONN.get(k_lock):
+        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="too many attempts, try later")
+
+    stored = REDIS_CONN.get(k_code)
+    if not stored:
+        return get_json_result(data=False, code=settings.RetCode.NOT_EFFECTIVE, message="expired otp")
+
+    try:
+        stored_hash, salt_hex = str(stored).split(":", 1)
+        salt = bytes.fromhex(salt_hex)
+    except Exception:
+        return get_json_result(data=False, code=settings.RetCode.EXCEPTION_ERROR, message="otp storage corrupted")
+
+    # Case-insensitive verification: OTP generated uppercase
+    calc = hash_code(otp.upper(), salt)
+    if calc != stored_hash:
+        # bump attempts
+        try:
+            attempts = int(REDIS_CONN.get(k_attempts) or 0) + 1
+        except Exception:
+            attempts = 1
+        REDIS_CONN.set(k_attempts, attempts, OTP_TTL_SECONDS)
+        if attempts >= ATTEMPT_LIMIT:
+            REDIS_CONN.set(k_lock, int(time.time()), ATTEMPT_LOCK_SECONDS)
+        return get_json_result(data=False, code=settings.RetCode.AUTHENTICATION_ERROR, message="expired otp")
+
+    # Success: consume OTP and reset password
+    REDIS_CONN.delete(k_code)
+    REDIS_CONN.delete(k_attempts)
+    REDIS_CONN.delete(k_last)
+    REDIS_CONN.delete(k_lock)
+
+    try:
+        UserService.update_user_password(user.id, new_pwd)
+    except Exception as e:
+        logging.exception(e)
+        return get_json_result(data=False, code=settings.RetCode.EXCEPTION_ERROR, message="failed to reset password")
+
+    # Auto login (reuse login flow)
+    user.access_token = get_uuid()
+    login_user(user)
+    user.update_time = (current_timestamp(),)
+    user.update_date = (datetime_format(datetime.now()),)
+    user.save()
+    msg = "Password reset successful. Logged in."
+    return construct_response(data=user.to_json(), auth=user.get_id(), message=msg)

api/common/README.md

@@ -0,0 +1,2 @@
+The python files in this directory are shared between service. They contain common utilities, models, and functions that can be used across various
+services to ensure consistency and reduce code duplication.

api/common/base64.py

@@ -0,0 +1,21 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import base64
+
+def encode_to_base64(input_string):
+    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
+    return base64_encoded.decode('utf-8')

api/common/check_team_permission.py

@@ -0,0 +1,59 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
+from api.db import TenantPermission
+from api.db.db_models import File, Knowledgebase
+from api.db.services.file_service import FileService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.user_service import TenantService
+
+
+def check_kb_team_permission(kb: dict | Knowledgebase, other: str) -> bool:
+    kb = kb.to_dict() if isinstance(kb, Knowledgebase) else kb
+
+    kb_tenant_id = kb["tenant_id"]
+
+    if kb_tenant_id == other:
+        return True
+
+    if kb["permission"] != TenantPermission.TEAM:
+        return False
+
+    joined_tenants = TenantService.get_joined_tenants_by_user_id(other)
+    return any(tenant["tenant_id"] == kb_tenant_id for tenant in joined_tenants)
+
+
+def check_file_team_permission(file: dict | File, other: str) -> bool:
+    file = file.to_dict() if isinstance(file, File) else file
+
+    file_tenant_id = file["tenant_id"]
+    if file_tenant_id == other:
+        return True
+
+    file_id = file["id"]
+
+    kb_ids = [kb_info["kb_id"] for kb_info in FileService.get_kb_id_by_file_id(file_id)]
+
+    for kb_id in kb_ids:
+        ok, kb = KnowledgebaseService.get_by_id(kb_id)
+        if not ok:
+            continue
+
+        if check_kb_team_permission(kb, other):
+            return True
+
+    return False

api/common/exceptions.py

@@ -0,0 +1,43 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+
+class AdminException(Exception):
+    def __init__(self, message, code=400):
+        super().__init__(message)
+        self.type = "admin"
+        self.code = code
+        self.message = message
+
+
+class UserNotFoundError(AdminException):
+    def __init__(self, username):
+        super().__init__(f"User '{username}' not found", 404)
+
+
+class UserAlreadyExistsError(AdminException):
+    def __init__(self, username):
+        super().__init__(f"User '{username}' already exists", 409)
+
+
+class CannotDeleteAdminError(AdminException):
+    def __init__(self):
+        super().__init__("Cannot delete admin account", 403)
+
+
+class NotAdminError(AdminException):
+    def __init__(self, username):
+        super().__init__(f"User '{username}' is not admin", 403)

api/db/__init__.py

@@ -127,4 +127,15 @@ class MCPServerType(StrEnum):
 VALID_MCP_SERVER_TYPES = {MCPServerType.SSE, MCPServerType.STREAMABLE_HTTP}
 
 
+class PipelineTaskType(StrEnum):
+    PARSE = "Parse"
+    DOWNLOAD = "Download"
+    RAPTOR = "RAPTOR"
+    GRAPH_RAG = "GraphRAG"
+    MINDMAP = "Mindmap"
+
+
+VALID_PIPELINE_TASK_TYPES = {PipelineTaskType.PARSE, PipelineTaskType.DOWNLOAD, PipelineTaskType.RAPTOR, PipelineTaskType.GRAPH_RAG, PipelineTaskType.MINDMAP}
+
+
 KNOWLEDGEBASE_FOLDER_NAME=".knowledgebase"

api/db/db_models.py

@@ -313,9 +313,75 @@ class RetryingPooledMySQLDatabase(PooledMySQLDatabase):
                     raise
 
 
+class RetryingPooledPostgresqlDatabase(PooledPostgresqlDatabase):
+    def __init__(self, *args, **kwargs):
+        self.max_retries = kwargs.pop("max_retries", 5)
+        self.retry_delay = kwargs.pop("retry_delay", 1)
+        super().__init__(*args, **kwargs)
+
+    def execute_sql(self, sql, params=None, commit=True):
+        for attempt in range(self.max_retries + 1):
+            try:
+                return super().execute_sql(sql, params, commit)
+            except (OperationalError, InterfaceError) as e:
+                # PostgreSQL specific error codes
+                # 57P01: admin_shutdown
+                # 57P02: crash_shutdown
+                # 57P03: cannot_connect_now
+                # 08006: connection_failure
+                # 08003: connection_does_not_exist
+                # 08000: connection_exception
+                error_messages = ['connection', 'server closed', 'connection refused', 
+                                'no connection to the server', 'terminating connection']
+                
+                should_retry = any(msg in str(e).lower() for msg in error_messages)
+
+                if should_retry and attempt < self.max_retries:
+                    logging.warning(
+                        f"PostgreSQL connection issue (attempt {attempt+1}/{self.max_retries}): {e}"
+                    )
+                    self._handle_connection_loss()
+                    time.sleep(self.retry_delay * (2 ** attempt))
+                else:
+                    logging.error(f"PostgreSQL execution failure: {e}")
+                    raise
+        return None
+
+    def _handle_connection_loss(self):
+        try:
+            self.close()
+        except Exception:
+            pass
+        try:
+            self.connect()
+        except Exception as e:
+            logging.error(f"Failed to reconnect to PostgreSQL: {e}")
+            time.sleep(0.1)
+            self.connect()
+
+    def begin(self):
+        for attempt in range(self.max_retries + 1):
+            try:
+                return super().begin()
+            except (OperationalError, InterfaceError) as e:
+                error_messages = ['connection', 'server closed', 'connection refused',
+                                'no connection to the server', 'terminating connection']
+                
+                should_retry = any(msg in str(e).lower() for msg in error_messages)
+
+                if should_retry and attempt < self.max_retries:
+                    logging.warning(
+                        f"PostgreSQL connection lost during transaction (attempt {attempt+1}/{self.max_retries})"
+                    )
+                    self._handle_connection_loss()
+                    time.sleep(self.retry_delay * (2 ** attempt))
+                else:
+                    raise
+
+
 class PooledDatabase(Enum):
     MYSQL = RetryingPooledMySQLDatabase
-    POSTGRES = PooledPostgresqlDatabase
+    POSTGRES = RetryingPooledPostgresqlDatabase
 
 
 class DatabaseMigrator(Enum):
@@ -641,7 +707,7 @@ class TenantLLM(DataBaseModel):
     llm_factory = CharField(max_length=128, null=False, help_text="LLM factory name", index=True)
     model_type = CharField(max_length=128, null=True, help_text="LLM, Text Embedding, Image2Text, ASR", index=True)
     llm_name = CharField(max_length=128, null=True, help_text="LLM name", default="", index=True)
-    api_key = CharField(max_length=2048, null=True, help_text="API KEY", index=True)
+    api_key = TextField(null=True, help_text="API KEY")
     api_base = CharField(max_length=255, null=True, help_text="API Base")
     max_tokens = IntegerField(default=8192, index=True)
     used_tokens = IntegerField(default=0, index=True)
@@ -684,8 +750,17 @@ class Knowledgebase(DataBaseModel):
     vector_similarity_weight = FloatField(default=0.3, index=True)
 
     parser_id = CharField(max_length=32, null=False, help_text="default parser ID", default=ParserType.NAIVE.value, index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)
     parser_config = JSONField(null=False, default={"pages": [[1, 1000000]]})
     pagerank = IntegerField(default=0, index=False)
+
+    graphrag_task_id = CharField(max_length=32, null=True, help_text="Graph RAG task ID", index=True)
+    graphrag_task_finish_at = DateTimeField(null=True)
+    raptor_task_id = CharField(max_length=32, null=True, help_text="RAPTOR task ID", index=True)
+    raptor_task_finish_at = DateTimeField(null=True)
+    mindmap_task_id = CharField(max_length=32, null=True, help_text="Mindmap task ID", index=True)
+    mindmap_task_finish_at = DateTimeField(null=True)
+
     status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
 
     def __str__(self):
@@ -700,6 +775,7 @@ class Document(DataBaseModel):
     thumbnail = TextField(null=True, help_text="thumbnail base64 string")
     kb_id = CharField(max_length=256, null=False, index=True)
     parser_id = CharField(max_length=32, null=False, help_text="default parser ID", index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="pipleline ID", index=True)
     parser_config = JSONField(null=False, default={"pages": [[1, 1000000]]})
     source_type = CharField(max_length=128, null=False, default="local", help_text="where dose this document come from", index=True)
     type = CharField(max_length=32, null=False, help_text="file extension", index=True)
@@ -942,6 +1018,32 @@ class Search(DataBaseModel):
         db_table = "search"
 
 
+class PipelineOperationLog(DataBaseModel):
+    id = CharField(max_length=32, primary_key=True)
+    document_id = CharField(max_length=32, index=True)
+    tenant_id = CharField(max_length=32, null=False, index=True)
+    kb_id = CharField(max_length=32, null=False, index=True)
+    pipeline_id = CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)
+    pipeline_title = CharField(max_length=32, null=True, help_text="Pipeline title", index=True)
+    parser_id = CharField(max_length=32, null=False, help_text="Parser ID", index=True)
+    document_name = CharField(max_length=255, null=False, help_text="File name")
+    document_suffix = CharField(max_length=255, null=False, help_text="File suffix")
+    document_type = CharField(max_length=255, null=False, help_text="Document type")
+    source_from = CharField(max_length=255, null=False, help_text="Source")
+    progress = FloatField(default=0, index=True)
+    progress_msg = TextField(null=True, help_text="process message", default="")
+    process_begin_at = DateTimeField(null=True, index=True)
+    process_duration = FloatField(default=0)
+    dsl = JSONField(null=True, default=dict)
+    task_type = CharField(max_length=32, null=False, default="")
+    operation_status = CharField(max_length=32, null=False, help_text="Operation status")
+    avatar = TextField(null=True, help_text="avatar base64 string")
+    status = CharField(max_length=1, null=True, help_text="is it validate(0: wasted, 1: validate)", default="1", index=True)
+
+    class Meta:
+        db_table = "pipeline_operation_log"
+
+
 def migrate_db():
     logging.disable(logging.ERROR)
     migrator = DatabaseMigrator[settings.DATABASE_TYPE.upper()].value(DB)
@@ -1058,7 +1160,6 @@ def migrate_db():
         migrate(migrator.add_column("dialog", "meta_data_filter", JSONField(null=True, default={})))
     except Exception:
         pass
-
     try:
         migrate(migrator.alter_column_type("canvas_template", "title", JSONField(null=True, default=dict, help_text="Canvas title")))
     except Exception:
@@ -1075,4 +1176,40 @@ def migrate_db():
         migrate(migrator.add_column("canvas_template", "canvas_category", CharField(max_length=32, null=False, default="agent_canvas", help_text="agent_canvas|dataflow_canvas", index=True)))
     except Exception:
         pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "pipeline_id", CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("document", "pipeline_id", CharField(max_length=32, null=True, help_text="Pipeline ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "graphrag_task_id", CharField(max_length=32, null=True, help_text="Gragh RAG task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "raptor_task_id", CharField(max_length=32, null=True, help_text="RAPTOR task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "graphrag_task_finish_at", DateTimeField(null=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "raptor_task_finish_at", CharField(null=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "mindmap_task_id", CharField(max_length=32, null=True, help_text="Mindmap task ID", index=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.add_column("knowledgebase", "mindmap_task_finish_at", CharField(null=True)))
+    except Exception:
+        pass
+    try:
+        migrate(migrator.alter_column_type("tenant_llm", "api_key", TextField(null=True, help_text="API KEY")))
+    except Exception:
+        pass
     logging.disable(logging.NOTSET)

api/db/init_data.py

@@ -14,7 +14,6 @@
 #  limitations under the License.
 #
 import logging
-import base64
 import json
 import os
 import time
@@ -32,11 +31,7 @@ from api.db.services.llm_service import LLMService, LLMBundle, get_init_tenant_l
 from api.db.services.user_service import TenantService, UserTenantService
 from api import settings
 from api.utils.file_utils import get_project_base_directory
-
+from api.common.base64 import encode_to_base64
-
-def encode_to_base64(input_string):
-    base64_encoded = base64.b64encode(input_string.encode('utf-8'))
-    return base64_encoded.decode('utf-8')
 
 
 def init_superuser():

api/db/joint_services/user_account_service.py

@@ -17,13 +17,26 @@ import logging
 import uuid
 
 from api import settings
-from api.db import FileType, UserTenantRole
+from api.utils.api_utils import group_by
-from api.db.db_models import TenantLLM
+from api.db import FileType, UserTenantRole, ActiveEnum
+from api.db.services.api_service import APITokenService, API4ConversationService
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.conversation_service import ConversationService
+from api.db.services.dialog_service import DialogService
+from api.db.services.document_service import DocumentService
+from api.db.services.file2document_service import File2DocumentService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.langfuse_service import TenantLangfuseService
 from api.db.services.llm_service import get_init_tenant_llm
 from api.db.services.file_service import FileService
+from api.db.services.mcp_server_service import MCPServerService
+from api.db.services.search_service import SearchService
+from api.db.services.task_service import TaskService
 from api.db.services.tenant_llm_service import TenantLLMService
+from api.db.services.user_canvas_version import UserCanvasVersionService
 from api.db.services.user_service import TenantService, UserService, UserTenantService
-
+from rag.utils.storage_factory import STORAGE_IMPL
+from rag.nlp import search
 
 
 def create_new_user(user_info: dict) -> dict:
@@ -104,7 +117,7 @@ def create_new_user(user_info: dict) -> dict:
         except Exception as e:
             logging.exception(e)
         try:
-            TenantLLM.delete().where(TenantLLM.tenant_id == user_id).execute()
+            TenantLLMService.delete_by_tenant_id(user_id)
         except Exception as e:
             logging.exception(e)
         try:
@@ -118,3 +131,197 @@ def create_new_user(user_info: dict) -> dict:
             logging.exception(e)
         # reraise
         raise create_error
+
+
+def delete_user_data(user_id: str) -> dict:
+    # use user_id to delete
+    usr = UserService.filter_by_id(user_id)
+    if not usr:
+        return {"success": False, "message": f"{user_id} can't be found."}
+    # check is inactive and not admin
+    if usr.is_active == ActiveEnum.ACTIVE.value:
+        return {"success": False, "message": f"{user_id} is active and can't be deleted."}
+    if usr.is_superuser:
+        return {"success": False, "message": "Can't delete the super user."}
+    # tenant info
+    tenants = UserTenantService.get_user_tenant_relation_by_user_id(usr.id)
+    owned_tenant = [t for t in tenants if t["role"] == UserTenantRole.OWNER.value]
+
+    done_msg = ''
+    try:
+        # step1. delete owned tenant info
+        if owned_tenant:
+            done_msg += "Start to delete owned tenant.\n"
+            tenant_id = owned_tenant[0]["tenant_id"]
+            kb_ids = KnowledgebaseService.get_kb_ids(usr.id)
+            # step1.1 delete knowledgebase related file and info
+            if kb_ids:
+                # step1.1.1 delete files in storage, remove bucket
+                for kb_id in kb_ids:
+                    if STORAGE_IMPL.bucket_exists(kb_id):
+                        STORAGE_IMPL.remove_bucket(kb_id)
+                done_msg += f"- Removed {len(kb_ids)} dataset's buckets.\n"
+                # step1.1.2 delete file and document info in db
+                doc_ids = DocumentService.get_all_doc_ids_by_kb_ids(kb_ids)
+                if doc_ids:
+                    doc_delete_res = DocumentService.delete_by_ids([i["id"] for i in doc_ids])
+                    done_msg += f"- Deleted {doc_delete_res} document records.\n"
+                    task_delete_res = TaskService.delete_by_doc_ids([i["id"] for i in doc_ids])
+                    done_msg += f"- Deleted {task_delete_res} task records.\n"
+                file_ids = FileService.get_all_file_ids_by_tenant_id(usr.id)
+                if file_ids:
+                    file_delete_res = FileService.delete_by_ids([f["id"] for f in file_ids])
+                    done_msg += f"- Deleted {file_delete_res} file records.\n"
+                if doc_ids or file_ids:
+                    file2doc_delete_res = File2DocumentService.delete_by_document_ids_or_file_ids(
+                        [i["id"] for i in doc_ids],
+                        [f["id"] for f in file_ids]
+                    )
+                    done_msg += f"- Deleted {file2doc_delete_res} document-file relation records.\n"
+                # step1.1.3 delete chunk in es
+                r = settings.docStoreConn.delete({"kb_id": kb_ids},
+                                         search.index_name(tenant_id), kb_ids)
+                done_msg += f"- Deleted {r} chunk records.\n"
+                kb_delete_res = KnowledgebaseService.delete_by_ids(kb_ids)
+                done_msg += f"- Deleted {kb_delete_res} knowledgebase records.\n"
+                # step1.1.4 delete agents
+                agent_delete_res = delete_user_agents(usr.id)
+                done_msg += f"- Deleted {agent_delete_res['agents_deleted_count']} agent, {agent_delete_res['version_deleted_count']} versions records.\n"
+                # step1.1.5 delete dialogs
+                dialog_delete_res = delete_user_dialogs(usr.id)
+                done_msg += f"- Deleted {dialog_delete_res['dialogs_deleted_count']} dialogs, {dialog_delete_res['conversations_deleted_count']} conversations, {dialog_delete_res['api_token_deleted_count']} api tokens, {dialog_delete_res['api4conversation_deleted_count']} api4conversations.\n"
+                # step1.1.6 delete mcp server
+                mcp_delete_res = MCPServerService.delete_by_tenant_id(usr.id)
+                done_msg += f"- Deleted {mcp_delete_res} MCP server.\n"
+                # step1.1.7 delete search
+                search_delete_res = SearchService.delete_by_tenant_id(usr.id)
+                done_msg += f"- Deleted {search_delete_res} search records.\n"
+            # step1.2 delete tenant_llm and tenant_langfuse
+            llm_delete_res = TenantLLMService.delete_by_tenant_id(tenant_id)
+            done_msg += f"- Deleted {llm_delete_res} tenant-LLM records.\n"
+            langfuse_delete_res = TenantLangfuseService.delete_ty_tenant_id(tenant_id)
+            done_msg += f"- Deleted {langfuse_delete_res} langfuse records.\n"
+            # step1.3 delete own tenant
+            tenant_delete_res = TenantService.delete_by_id(tenant_id)
+            done_msg += f"- Deleted {tenant_delete_res} tenant.\n"
+        # step2 delete user-tenant relation
+        if tenants:
+            # step2.1 delete docs and files in joined team
+            joined_tenants = [t for t in tenants if t["role"] == UserTenantRole.NORMAL.value]
+            if joined_tenants:
+                done_msg += "Start to delete data in joined tenants.\n"
+                created_documents = DocumentService.get_all_docs_by_creator_id(usr.id)
+                if created_documents:
+                    # step2.1.1 delete files
+                    doc_file_info = File2DocumentService.get_by_document_ids([d['id'] for d in created_documents])
+                    created_files = FileService.get_by_ids([f['file_id'] for f in doc_file_info])
+                    if created_files:
+                        # step2.1.1.1 delete file in storage
+                        for f in created_files:
+                            STORAGE_IMPL.rm(f.parent_id, f.location)
+                        done_msg += f"- Deleted {len(created_files)} uploaded file.\n"
+                        # step2.1.1.2 delete file record
+                        file_delete_res = FileService.delete_by_ids([f.id for f in created_files])
+                        done_msg += f"- Deleted {file_delete_res} file records.\n"
+                    # step2.1.2 delete document-file relation record
+                    file2doc_delete_res = File2DocumentService.delete_by_document_ids_or_file_ids(
+                        [d['id'] for d in created_documents],
+                        [f.id for f in created_files]
+                    )
+                    done_msg += f"- Deleted {file2doc_delete_res} document-file relation records.\n"
+                    # step2.1.3 delete chunks
+                    doc_groups = group_by(created_documents, "tenant_id")
+                    kb_grouped_doc = {k: group_by(v, "kb_id") for k, v in doc_groups.items()}
+                    # chunks in {'tenant_id': {'kb_id': [{'id': doc_id}]}} structure
+                    chunk_delete_res = 0
+                    kb_doc_info = {}
+                    for _tenant_id, kb_doc in kb_grouped_doc.items():
+                        for _kb_id, docs in kb_doc.items():
+                            chunk_delete_res += settings.docStoreConn.delete(
+                                {"doc_id": [d["id"] for d in docs]},
+                                search.index_name(_tenant_id), _kb_id
+                            )
+                            # record doc info
+                            if _kb_id in kb_doc_info.keys():
+                                kb_doc_info[_kb_id]['doc_num'] += 1
+                                kb_doc_info[_kb_id]['token_num'] += sum([d["token_num"] for d in docs])
+                                kb_doc_info[_kb_id]['chunk_num'] += sum([d["chunk_num"] for d in docs])
+                            else:
+                                kb_doc_info[_kb_id] = {
+                                    'doc_num': 1,
+                                    'token_num': sum([d["token_num"] for d in docs]),
+                                    'chunk_num': sum([d["chunk_num"] for d in docs])
+                                }
+                    done_msg += f"- Deleted {chunk_delete_res} chunks.\n"
+                    # step2.1.4 delete tasks
+                    task_delete_res = TaskService.delete_by_doc_ids([d['id'] for d in created_documents])
+                    done_msg += f"- Deleted {task_delete_res} tasks.\n"
+                    # step2.1.5 delete document record
+                    doc_delete_res = DocumentService.delete_by_ids([d['id'] for d in created_documents])
+                    done_msg += f"- Deleted {doc_delete_res} documents.\n"
+                    # step2.1.6 update knowledge base doc&chunk&token cnt
+                    for kb_id, doc_num in kb_doc_info.items():
+                        KnowledgebaseService.decrease_document_num_in_delete(kb_id, doc_num)
+
+            # step2.2 delete relation
+            user_tenant_delete_res = UserTenantService.delete_by_ids([t["id"] for t in tenants])
+            done_msg += f"- Deleted {user_tenant_delete_res} user-tenant records.\n"
+        # step3 finally delete user
+        user_delete_res = UserService.delete_by_id(usr.id)
+        done_msg += f"- Deleted {user_delete_res} user.\nDelete done!"
+
+        return {"success": True, "message": f"Successfully deleted user. Details:\n{done_msg}"}
+
+    except Exception as e:
+        logging.exception(e)
+        return {"success": False, "message": f"Error: {str(e)}. Already done:\n{done_msg}"}
+
+
+def delete_user_agents(user_id: str) -> dict:
+    """
+    use user_id to delete
+    :return: {
+        "agents_deleted_count": 1,
+        "version_deleted_count": 2
+    }
+    """
+    agents_deleted_count, agents_version_deleted_count = 0, 0
+    user_agents = UserCanvasService.get_all_agents_by_tenant_ids([user_id], user_id)
+    if user_agents:
+        agents_version = UserCanvasVersionService.get_all_canvas_version_by_canvas_ids([a['id'] for a in user_agents])
+        agents_version_deleted_count = UserCanvasVersionService.delete_by_ids([v['id'] for v in agents_version])
+        agents_deleted_count = UserCanvasService.delete_by_ids([a['id'] for a in user_agents])
+    return {
+        "agents_deleted_count": agents_deleted_count,
+        "version_deleted_count": agents_version_deleted_count
+    }
+
+
+def delete_user_dialogs(user_id: str) -> dict:
+    """
+    use user_id to delete
+    :return: {
+        "dialogs_deleted_count": 1,
+        "conversations_deleted_count": 1,
+        "api_token_deleted_count": 2,
+        "api4conversation_deleted_count": 2
+    }
+    """
+    dialog_deleted_count, conversations_deleted_count, api_token_deleted_count, api4conversation_deleted_count = 0, 0, 0, 0
+    user_dialogs = DialogService.get_all_dialogs_by_tenant_id(user_id)
+    if user_dialogs:
+        # delete conversation
+        conversations = ConversationService.get_all_conversation_by_dialog_ids([ud['id'] for ud in user_dialogs])
+        conversations_deleted_count = ConversationService.delete_by_ids([c['id'] for c in conversations])
+        # delete api token
+        api_token_deleted_count = APITokenService.delete_by_tenant_id(user_id)
+        # delete api for conversation
+        api4conversation_deleted_count = API4ConversationService.delete_by_dialog_ids([ud['id'] for ud in user_dialogs])
+        # delete dialog at last
+        dialog_deleted_count = DialogService.delete_by_ids([ud['id'] for ud in user_dialogs])
+    return {
+        "dialogs_deleted_count": dialog_deleted_count,
+        "conversations_deleted_count": conversations_deleted_count,
+        "api_token_deleted_count": api_token_deleted_count,
+        "api4conversation_deleted_count": api4conversation_deleted_count
+    }

api/db/services/api_service.py

@@ -35,6 +35,11 @@ class APITokenService(CommonService):
             cls.model.token == token
         )
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
 
 class API4ConversationService(CommonService):
     model = API4Conversation
@@ -100,3 +105,8 @@ class API4ConversationService(CommonService):
             cls.model.create_date <= to_date,
             cls.model.source == source
         ).group_by(cls.model.create_date.truncate("day")).dicts()
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_dialog_ids(cls, dialog_ids):
+        return cls.model.delete().where(cls.model.dialog_id.in_(dialog_ids)).execute()

api/db/services/canvas_service.py

@@ -66,6 +66,7 @@ class UserCanvasService(CommonService):
     def get_all_agents_by_tenant_ids(cls, tenant_ids, user_id):
         # will get all permitted agents, be cautious
         fields = [
+            cls.model.id,
             cls.model.title,
             cls.model.permission,
             cls.model.canvas_type,
@@ -93,7 +94,7 @@ class UserCanvasService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_by_tenant_id(cls, pid):
+    def get_by_canvas_id(cls, pid):
         try:
 
             fields = [
@@ -125,7 +126,7 @@ class UserCanvasService(CommonService):
     @DB.connection_context()
     def get_by_tenant_ids(cls, joined_tenant_ids, user_id,
                           page_number, items_per_page,
-                          orderby, desc, keywords, canvas_category=CanvasCategory.Agent,
+                          orderby, desc, keywords, canvas_category=None
                           ):
         fields = [
             cls.model.id,
@@ -134,6 +135,7 @@ class UserCanvasService(CommonService):
             cls.model.dsl,
             cls.model.description,
             cls.model.permission,
+            cls.model.user_id.alias("tenant_id"),
             User.nickname,
             User.avatar.alias('tenant_avatar'),
             cls.model.update_time,
@@ -141,31 +143,30 @@ class UserCanvasService(CommonService):
         ]
         if keywords:
             agents = cls.model.select(*fields).join(User, on=(cls.model.user_id == User.id)).where(
-                ((cls.model.user_id.in_(joined_tenant_ids) & (cls.model.permission ==
+                (((cls.model.user_id.in_(joined_tenant_ids)) & (cls.model.permission == TenantPermission.TEAM.value)) | (cls.model.user_id == user_id)),
-                                                                TenantPermission.TEAM.value)) | (
-                    cls.model.user_id == user_id)),
                 (fn.LOWER(cls.model.title).contains(keywords.lower()))
             )
         else:
             agents = cls.model.select(*fields).join(User, on=(cls.model.user_id == User.id)).where(
-                ((cls.model.user_id.in_(joined_tenant_ids) & (cls.model.permission ==
+                (((cls.model.user_id.in_(joined_tenant_ids)) & (cls.model.permission == TenantPermission.TEAM.value)) | (cls.model.user_id == user_id))
-                                                                TenantPermission.TEAM.value)) | (
-                    cls.model.user_id == user_id))
             )
-        agents = agents.where(cls.model.canvas_category == canvas_category)
+        if canvas_category:
+            agents = agents.where(cls.model.canvas_category == canvas_category)
         if desc:
             agents = agents.order_by(cls.model.getter_by(orderby).desc())
         else:
             agents = agents.order_by(cls.model.getter_by(orderby).asc())
+
         count = agents.count()
-        agents = agents.paginate(page_number, items_per_page)
+        if page_number and items_per_page:
+            agents = agents.paginate(page_number, items_per_page)
         return list(agents.dicts()), count
 
     @classmethod
     @DB.connection_context()
     def accessible(cls, canvas_id, tenant_id):
         from api.db.services.user_service import UserTenantService
-        e, c = UserCanvasService.get_by_tenant_id(canvas_id)
+        e, c = UserCanvasService.get_by_canvas_id(canvas_id)
         if not e:
             return False
 

api/db/services/conversation_service.py

@@ -48,6 +48,21 @@ class ConversationService(CommonService):
 
         return list(sessions.dicts())
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_conversation_by_dialog_ids(cls, dialog_ids):
+        sessions = cls.model.select().where(cls.model.dialog_id.in_(dialog_ids))
+        sessions.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            s_batch = sessions.offset(offset).limit(limit)
+            _temp = list(s_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
 
 def structure_answer(conv, ans, message_id, session_id):
     reference = ans["reference"]

api/db/services/dialog_service.py

@@ -159,6 +159,22 @@ class DialogService(CommonService):
 
         return list(dialogs.dicts()), count
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_dialogs_by_tenant_id(cls, tenant_id):
+        fields = [cls.model.id]
+        dialogs = cls.model.select(*fields).where(cls.model.tenant_id == tenant_id)
+        dialogs.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            d_batch = dialogs.offset(offset).limit(limit)
+            _temp = list(d_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
 
 def chat_solo(dialog, messages, stream=True):
     if TenantLLMService.llm_id2llm_type(dialog.llm_id) == "image2text":
@@ -354,7 +370,7 @@ def chat(dialog, messages, stream=True, **kwargs):
         chat_mdl.bind_tools(toolcall_session, tools)
     bind_models_ts = timer()
 
-    retriever = settings.retrievaler
+    retriever = settings.retriever
     questions = [m["content"] for m in messages if m["role"] == "user"][-3:]
     attachments = kwargs["doc_ids"].split(",") if "doc_ids" in kwargs else []
     if "doc_ids" in messages[-1]:
@@ -450,13 +466,17 @@ def chat(dialog, messages, stream=True, **kwargs):
                     rerank_mdl=rerank_mdl,
                     rank_feature=label_question(" ".join(questions), kbs),
                 )
+                if prompt_config.get("toc_enhance"):
+                    cks = retriever.retrieval_by_toc(" ".join(questions), kbinfos["chunks"], tenant_ids, chat_mdl, dialog.top_n)
+                    if cks:
+                        kbinfos["chunks"] = cks
             if prompt_config.get("tavily_api_key"):
                 tav = Tavily(prompt_config["tavily_api_key"])
                 tav_res = tav.retrieve_chunks(" ".join(questions))
                 kbinfos["chunks"].extend(tav_res["chunks"])
                 kbinfos["doc_aggs"].extend(tav_res["doc_aggs"])
             if prompt_config.get("use_kg"):
-                ck = settings.kg_retrievaler.retrieval(" ".join(questions), tenant_ids, dialog.kb_ids, embd_mdl,
+                ck = settings.kg_retriever.retrieval(" ".join(questions), tenant_ids, dialog.kb_ids, embd_mdl,
                                                        LLMBundle(dialog.tenant_id, LLMType.CHAT))
                 if ck["content_with_weight"]:
                     kbinfos["chunks"].insert(0, ck)
@@ -642,7 +662,7 @@ Please write the SQL, only SQL, without any other explanations or text.
 
         logging.debug(f"{question} get SQL(refined): {sql}")
         tried_times += 1
-        return settings.retrievaler.sql_retrieval(sql, format="json"), sql
+        return settings.retriever.sql_retrieval(sql, format="json"), sql
 
     tbl, sql = get_table()
     if tbl is None:
@@ -736,7 +756,7 @@ def ask(question, kb_ids, tenant_id, chat_llm_name=None, search_config={}):
     embedding_list = list(set([kb.embd_id for kb in kbs]))
 
     is_knowledge_graph = all([kb.parser_id == ParserType.KG for kb in kbs])
-    retriever = settings.retrievaler if not is_knowledge_graph else settings.kg_retrievaler
+    retriever = settings.retriever if not is_knowledge_graph else settings.kg_retriever
 
     embd_mdl = LLMBundle(tenant_id, LLMType.EMBEDDING, embedding_list[0])
     chat_mdl = LLMBundle(tenant_id, LLMType.CHAT, chat_llm_name)
@@ -832,7 +852,7 @@ def gen_mindmap(question, kb_ids, tenant_id, search_config={}):
             if not doc_ids:
                 doc_ids = None
 
-    ranks = settings.retrievaler.retrieval(
+    ranks = settings.retriever.retrieval(
         question=question,
         embd_mdl=embd_mdl,
         tenant_ids=tenant_ids,

api/db/services/document_service.py

@@ -24,12 +24,13 @@ from io import BytesIO
 
 import trio
 import xxhash
-from peewee import fn, Case
+from peewee import fn, Case, JOIN
 
 from api import settings
 from api.constants import IMG_BASE64_PREFIX, FILE_NAME_LEN_LIMIT
-from api.db import FileType, LLMType, ParserType, StatusEnum, TaskStatus, UserTenantRole
+from api.db import FileType, LLMType, ParserType, StatusEnum, TaskStatus, UserTenantRole, CanvasCategory
-from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File
+from api.db.db_models import DB, Document, Knowledgebase, Task, Tenant, UserTenant, File2Document, File, UserCanvas, \
+    User
 from api.db.db_utils import bulk_insert_into_db
 from api.db.services.common_service import CommonService
 from api.db.services.knowledgebase_service import KnowledgebaseService
@@ -51,6 +52,7 @@ class DocumentService(CommonService):
             cls.model.thumbnail,
             cls.model.kb_id,
             cls.model.parser_id,
+            cls.model.pipeline_id,
             cls.model.parser_config,
             cls.model.source_type,
             cls.model.type,
@@ -77,9 +79,12 @@ class DocumentService(CommonService):
     @classmethod
     @DB.connection_context()
     def get_list(cls, kb_id, page_number, items_per_page,
-                 orderby, desc, keywords, id, name):
+                 orderby, desc, keywords, id, name, suffix=None, run = None):
         fields = cls.get_cls_model_fields()
-        docs = cls.model.select(*fields).join(File2Document, on = (File2Document.document_id == cls.model.id)).join(File, on = (File.id == File2Document.file_id)).where(cls.model.kb_id == kb_id)
+        docs = cls.model.select(*[*fields, UserCanvas.title]).join(File2Document, on = (File2Document.document_id == cls.model.id))\
+            .join(File, on = (File.id == File2Document.file_id))\
+            .join(UserCanvas, on = ((cls.model.pipeline_id == UserCanvas.id) & (UserCanvas.canvas_category == CanvasCategory.DataFlow.value)), join_type=JOIN.LEFT_OUTER)\
+            .where(cls.model.kb_id == kb_id)
         if id:
             docs = docs.where(
                 cls.model.id == id)
@@ -91,6 +96,10 @@ class DocumentService(CommonService):
             docs = docs.where(
                 fn.LOWER(cls.model.name).contains(keywords.lower())
             )
+        if suffix:
+            docs = docs.where(cls.model.suffix.in_(suffix))
+        if run:
+            docs = docs.where(cls.model.run.in_(run))
         if desc:
             docs = docs.order_by(cls.model.getter_by(orderby).desc())
         else:
@@ -117,12 +126,22 @@ class DocumentService(CommonService):
                      orderby, desc, keywords, run_status, types, suffix):
         fields = cls.get_cls_model_fields()
         if keywords:
-            docs = cls.model.select(*fields).join(File2Document, on=(File2Document.document_id == cls.model.id)).join(File, on=(File.id == File2Document.file_id)).where(
+            docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name"), User.nickname])\
-                (cls.model.kb_id == kb_id),
+                .join(File2Document, on=(File2Document.document_id == cls.model.id))\
-                (fn.LOWER(cls.model.name).contains(keywords.lower()))
+                .join(File, on=(File.id == File2Document.file_id))\
-            )
+                .join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
+                .join(User, on=(cls.model.created_by == User.id), join_type=JOIN.LEFT_OUTER)\
+                .where(
+                    (cls.model.kb_id == kb_id),
+                    (fn.LOWER(cls.model.name).contains(keywords.lower()))
+                )
         else:
-            docs = cls.model.select(*fields).join(File2Document, on=(File2Document.document_id == cls.model.id)).join(File, on=(File.id == File2Document.file_id)).where(cls.model.kb_id == kb_id)
+            docs = cls.model.select(*[*fields, UserCanvas.title.alias("pipeline_name"), User.nickname])\
+                .join(File2Document, on=(File2Document.document_id == cls.model.id))\
+                .join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
+                .join(File, on=(File.id == File2Document.file_id))\
+                .join(User, on=(cls.model.created_by == User.id), join_type=JOIN.LEFT_OUTER)\
+                .where(cls.model.kb_id == kb_id)
 
         if run_status:
             docs = docs.where(cls.model.run.in_(run_status))
@@ -228,6 +247,46 @@ class DocumentService(CommonService):
 
         return int(query.scalar()) or 0
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_doc_ids_by_kb_ids(cls, kb_ids):
+        fields = [cls.model.id]
+        docs = cls.model.select(*fields).where(cls.model.kb_id.in_(kb_ids))
+        docs.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later
+        offset, limit = 0, 100
+        res = []
+        while True:
+            doc_batch = docs.offset(offset).limit(limit)
+            _temp = list(doc_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
+    @classmethod
+    @DB.connection_context()
+    def get_all_docs_by_creator_id(cls, creator_id):
+        fields = [
+            cls.model.id, cls.model.kb_id, cls.model.token_num, cls.model.chunk_num, Knowledgebase.tenant_id
+        ]
+        docs = cls.model.select(*fields).join(Knowledgebase, on=(Knowledgebase.id == cls.model.kb_id)).where(
+            cls.model.created_by == creator_id
+        )
+        docs.order_by(cls.model.create_time.asc())
+        # maybe cause slow query by deep paginate, optimize later
+        offset, limit = 0, 100
+        res = []
+        while True:
+            doc_batch = docs.offset(offset).limit(limit)
+            _temp = list(doc_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def insert(cls, doc):
@@ -330,8 +389,7 @@ class DocumentService(CommonService):
                                process_duration=cls.model.process_duration + duration).where(
             cls.model.id == doc_id).execute()
         if num == 0:
-            raise LookupError(
+            logging.warning("Document not found which is supposed to be there")
-                "Document not found which is supposed to be there")
         num = Knowledgebase.update(
             token_num=Knowledgebase.token_num +
                       token_num,
@@ -597,17 +655,33 @@ class DocumentService(CommonService):
     @DB.connection_context()
     def update_progress(cls):
         docs = cls.get_unfinished_docs()
+
+        cls._sync_progress(docs)
+
+
+    @classmethod
+    @DB.connection_context()
+    def update_progress_immediately(cls, docs:list[dict]):
+        if not docs:
+            return
+
+        cls._sync_progress(docs)
+
+
+    @classmethod
+    @DB.connection_context()
+    def _sync_progress(cls, docs:list[dict]):
+        from api.db.services.task_service import TaskService
+
         for d in docs:
             try:
-                tsks = Task.query(doc_id=d["id"], order_by=Task.create_time)
+                tsks = TaskService.query(doc_id=d["id"], order_by=Task.create_time)
                 if not tsks:
                     continue
                 msg = []
                 prg = 0
                 finished = True
                 bad = 0
-                has_raptor = False
-                has_graphrag = False
                 e, doc = DocumentService.get_by_id(d["id"])
                 status = doc.run  # TaskStatus.RUNNING.value
                 priority = 0
@@ -619,24 +693,14 @@ class DocumentService(CommonService):
                     prg += t.progress if t.progress >= 0 else 0
                     if t.progress_msg.strip():
                         msg.append(t.progress_msg)
-                    if t.task_type == "raptor":
-                        has_raptor = True
-                    elif t.task_type == "graphrag":
-                        has_graphrag = True
                     priority = max(priority, t.priority)
                 prg /= len(tsks)
                 if finished and bad:
                     prg = -1
                     status = TaskStatus.FAIL.value
                 elif finished:
-                    if (d["parser_config"].get("raptor") or {}).get("use_raptor") and not has_raptor:
+                    prg = 1
-                        queue_raptor_o_graphrag_tasks(d, "raptor", priority)
+                    status = TaskStatus.DONE.value
-                        prg = 0.98 * len(tsks) / (len(tsks) + 1)
-                    elif (d["parser_config"].get("graphrag") or {}).get("use_graphrag") and not has_graphrag:
-                        queue_raptor_o_graphrag_tasks(d, "graphrag", priority)
-                        prg = 0.98 * len(tsks) / (len(tsks) + 1)
-                    else:
-                        status = TaskStatus.DONE.value
 
                 msg = "\n".join(sorted(msg))
                 info = {
@@ -648,7 +712,7 @@ class DocumentService(CommonService):
                     info["progress"] = prg
                 if msg:
                     info["progress_msg"] = msg
-                    if msg.endswith("created task graphrag") or msg.endswith("created task raptor"):
+                    if msg.endswith("created task graphrag") or msg.endswith("created task raptor") or msg.endswith("created task mindmap"):
                         info["progress_msg"] += "\n%d tasks are ahead in the queue..."%get_queue_length(priority)
                 else:
                     info["progress_msg"] = "%d tasks are ahead in the queue..."%get_queue_length(priority)
@@ -729,21 +793,28 @@ class DocumentService(CommonService):
             "cancelled": int(cancelled),
         }
 
-def queue_raptor_o_graphrag_tasks(doc, ty, priority):
+def queue_raptor_o_graphrag_tasks(sample_doc_id, ty, priority, fake_doc_id="", doc_ids=[]):
-    chunking_config = DocumentService.get_chunking_config(doc["id"])
+    """
+    You can provide a fake_doc_id to bypass the restriction of tasks at the knowledgebase level.
+    Optionally, specify a list of doc_ids to determine which documents participate in the task.
+    """
+    assert ty in ["graphrag", "raptor", "mindmap"], "type should be graphrag, raptor or mindmap"
+
+    chunking_config = DocumentService.get_chunking_config(sample_doc_id["id"])
     hasher = xxhash.xxh64()
     for field in sorted(chunking_config.keys()):
         hasher.update(str(chunking_config[field]).encode("utf-8"))
 
     def new_task():
-        nonlocal doc
+        nonlocal sample_doc_id
         return {
             "id": get_uuid(),
-            "doc_id": doc["id"],
+            "doc_id": sample_doc_id["id"],
             "from_page": 100000000,
             "to_page": 100000000,
             "task_type": ty,
-            "progress_msg":  datetime.now().strftime("%H:%M:%S") + " created task " + ty
+            "progress_msg":  datetime.now().strftime("%H:%M:%S") + " created task " + ty,
+            "begin_at": datetime.now(),
         }
 
     task = new_task()
@@ -752,7 +823,12 @@ def queue_raptor_o_graphrag_tasks(doc, ty, priority):
     hasher.update(ty.encode("utf-8"))
     task["digest"] = hasher.hexdigest()
     bulk_insert_into_db(Task, [task], True)
+
+    task["doc_id"] = fake_doc_id
+    task["doc_ids"] = doc_ids
+    DocumentService.begin2parse(sample_doc_id["id"])
     assert REDIS_CONN.queue_product(get_svr_queue_name(priority), message=task), "Can't access Redis. Please check the Redis' status."
+    return task["id"]
 
 
 def get_queue_length(priority):

api/db/services/file2document_service.py

@@ -38,6 +38,12 @@ class File2DocumentService(CommonService):
         objs = cls.model.select().where(cls.model.document_id == document_id)
         return objs
 
+    @classmethod
+    @DB.connection_context()
+    def get_by_document_ids(cls, document_ids):
+        objs = cls.model.select().where(cls.model.document_id.in_(document_ids))
+        return list(objs.dicts())
+
     @classmethod
     @DB.connection_context()
     def insert(cls, obj):
@@ -50,6 +56,15 @@ class File2DocumentService(CommonService):
     def delete_by_file_id(cls, file_id):
         return cls.model.delete().where(cls.model.file_id == file_id).execute()
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_document_ids_or_file_ids(cls, document_ids, file_ids):
+        if not document_ids:
+            return cls.model.delete().where(cls.model.file_id.in_(file_ids)).execute()
+        elif not file_ids:
+            return cls.model.delete().where(cls.model.document_id.in_(document_ids)).execute()
+        return cls.model.delete().where(cls.model.document_id.in_(document_ids) | cls.model.file_id.in_(file_ids)).execute()
+
     @classmethod
     @DB.connection_context()
     def delete_by_document_id(cls, doc_id):

api/db/services/file_service.py

@@ -161,6 +161,23 @@ class FileService(CommonService):
             result_ids.append(folder_id)
         return result_ids
 
+    @classmethod
+    @DB.connection_context()
+    def get_all_file_ids_by_tenant_id(cls, tenant_id):
+        fields = [cls.model.id]
+        files = cls.model.select(*fields).where(cls.model.tenant_id == tenant_id)
+        files.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            file_batch = files.offset(offset).limit(limit)
+            _temp = list(file_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def create_folder(cls, file, parent_id, name, count):
@@ -440,6 +457,7 @@ class FileService(CommonService):
                     "id": doc_id,
                     "kb_id": kb.id,
                     "parser_id": self.get_parser(filetype, filename, kb.parser_id),
+                    "pipeline_id": kb.pipeline_id,
                     "parser_config": kb.parser_config,
                     "created_by": user_id,
                     "type": filetype,
@@ -458,6 +476,16 @@ class FileService(CommonService):
 
         return err, files
 
+    @classmethod
+    @DB.connection_context()
+    def list_all_files_by_parent_id(cls, parent_id):
+        try:
+            files = cls.model.select().where((cls.model.parent_id == parent_id) & (cls.model.id != parent_id))
+            return list(files)
+        except Exception:
+            logging.exception("list_by_parent_id failed")
+            raise RuntimeError("Database error (list_by_parent_id)!")
+
     @staticmethod
     def parse_docs(file_objs, user_id):
         exe = ThreadPoolExecutor(max_workers=12)
@@ -495,7 +523,7 @@ class FileService(CommonService):
             return ParserType.AUDIO.value
         if re.search(r"\.(ppt|pptx|pages)$", filename):
             return ParserType.PRESENTATION.value
-        if re.search(r"\.(eml)$", filename):
+        if re.search(r"\.(msg|eml)$", filename):
             return ParserType.EMAIL.value
         return default
 

api/db/services/knowledgebase_service.py

@@ -15,10 +15,10 @@
 #
 from datetime import datetime
 
-from peewee import fn
+from peewee import fn, JOIN
 
 from api.db import StatusEnum, TenantPermission
-from api.db.db_models import DB, Document, Knowledgebase, Tenant, User, UserTenant
+from api.db.db_models import DB, Document, Knowledgebase, User, UserTenant, UserCanvas
 from api.db.services.common_service import CommonService
 from api.utils import current_timestamp, datetime_format
 
@@ -260,20 +260,29 @@ class KnowledgebaseService(CommonService):
             cls.model.token_num,
             cls.model.chunk_num,
             cls.model.parser_id,
+            cls.model.pipeline_id,
+            UserCanvas.title.alias("pipeline_name"),
+            UserCanvas.avatar.alias("pipeline_avatar"),
             cls.model.parser_config,
             cls.model.pagerank,
+            cls.model.graphrag_task_id,
+            cls.model.graphrag_task_finish_at,
+            cls.model.raptor_task_id,
+            cls.model.raptor_task_finish_at,
+            cls.model.mindmap_task_id,
+            cls.model.mindmap_task_finish_at,
             cls.model.create_time,
             cls.model.update_time
             ]
-        kbs = cls.model.select(*fields).join(Tenant, on=(
+        kbs = cls.model.select(*fields)\
-            (Tenant.id == cls.model.tenant_id) & (Tenant.status == StatusEnum.VALID.value))).where(
+                .join(UserCanvas, on=(cls.model.pipeline_id == UserCanvas.id), join_type=JOIN.LEFT_OUTER)\
+            .where(
             (cls.model.id == kb_id),
             (cls.model.status == StatusEnum.VALID.value)
-        )
+        ).dicts()
         if not kbs:
             return
-        d = kbs[0].to_dict()
+        return kbs[0]
-        return d
 
     @classmethod
     @DB.connection_context()
@@ -370,6 +379,7 @@ class KnowledgebaseService(CommonService):
         #     name: Optional name filter
         # Returns:
         #     List of knowledge bases
+        #     Total count of knowledge bases
         kbs = cls.model.select()
         if id:
             kbs = kbs.where(cls.model.id == id)
@@ -381,14 +391,16 @@ class KnowledgebaseService(CommonService):
                 cls.model.tenant_id == user_id))
             & (cls.model.status == StatusEnum.VALID.value)
         )
+
         if desc:
             kbs = kbs.order_by(cls.model.getter_by(orderby).desc())
         else:
             kbs = kbs.order_by(cls.model.getter_by(orderby).asc())
 
+        total = kbs.count()
         kbs = kbs.paginate(page_number, items_per_page)
 
-        return list(kbs.dicts())
+        return list(kbs.dicts()), total
 
     @classmethod
     @DB.connection_context()
@@ -471,3 +483,17 @@ class KnowledgebaseService(CommonService):
             else:
                 raise e
 
+    @classmethod
+    @DB.connection_context()
+    def decrease_document_num_in_delete(cls, kb_id, doc_num_info: dict):
+        kb_row = cls.model.get_by_id(kb_id)
+        if not kb_row:
+            raise RuntimeError(f"kb_id {kb_id} does not exist")
+        update_dict = {
+            'doc_num': kb_row.doc_num - doc_num_info['doc_num'],
+            'chunk_num': kb_row.chunk_num - doc_num_info['chunk_num'],
+            'token_num': kb_row.token_num - doc_num_info['token_num'],
+            'update_time': current_timestamp(),
+            'update_date': datetime_format(datetime.now())
+        }
+        return cls.model.update(update_dict).where(cls.model.id == kb_id).execute()

api/db/services/langfuse_service.py

@@ -51,6 +51,11 @@ class TenantLangfuseService(CommonService):
         except peewee.DoesNotExist:
             return None
 
+    @classmethod
+    @DB.connection_context()
+    def delete_ty_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
     @classmethod
     def update_by_tenant(cls, tenant_id, langfuse_keys):
         langfuse_keys["update_time"] = current_timestamp()

api/db/services/llm_service.py

@@ -205,32 +205,31 @@ class LLMBundle(LLM4Tenant):
             return txt
 
         return txt[last_think_end + len("</think>") :]
-    
+
     @staticmethod
     def _clean_param(chat_partial, **kwargs):
         func = chat_partial.func
         sig = inspect.signature(func)
-        keyword_args = []
         support_var_args = False
-        for param in sig.parameters.values():
+        allowed_params = set()
-            if param.kind == inspect.Parameter.VAR_KEYWORD or param.kind == inspect.Parameter.VAR_POSITIONAL:
-                support_var_args = True
-            elif param.kind == inspect.Parameter.KEYWORD_ONLY:
-                keyword_args.append(param.name)
 
-        use_kwargs = kwargs
+        for param in sig.parameters.values():
-        if not support_var_args:
+            if param.kind == inspect.Parameter.VAR_KEYWORD:
-            use_kwargs = {k: v for k, v in kwargs.items() if k in keyword_args}
+                support_var_args = True
-        return use_kwargs
+            elif param.kind in (inspect.Parameter.POSITIONAL_OR_KEYWORD, inspect.Parameter.KEYWORD_ONLY):
-        
+                allowed_params.add(param.name)
+        if support_var_args:
+            return kwargs
+        else:
+            return {k: v for k, v in kwargs.items() if k in allowed_params}
     def chat(self, system: str, history: list, gen_conf: dict = {}, **kwargs) -> str:
         if self.langfuse:
             generation = self.langfuse.start_generation(trace_context=self.trace_context, name="chat", model=self.llm_name, input={"system": system, "history": history})
 
-        chat_partial = partial(self.mdl.chat, system, history, gen_conf)
+        chat_partial = partial(self.mdl.chat, system, history, gen_conf, **kwargs)
         if self.is_tools and self.mdl.is_tools:
-            chat_partial = partial(self.mdl.chat_with_tools, system, history, gen_conf)
+            chat_partial = partial(self.mdl.chat_with_tools, system, history, gen_conf, **kwargs)
-            
+
         use_kwargs = self._clean_param(chat_partial, **kwargs)
         txt, used_tokens = chat_partial(**use_kwargs)
         txt = self._remove_reasoning_content(txt)
@@ -266,7 +265,7 @@ class LLMBundle(LLM4Tenant):
                 break
 
             if txt.endswith("</think>"):
-                ans = ans.rstrip("</think>")
+                ans = ans[: -len("</think>")]
 
             if not self.verbose_tool_use:
                 txt = re.sub(r"<tool_call>.*?</tool_call>", "", txt, flags=re.DOTALL)

api/db/services/mcp_server_service.py

@@ -33,7 +33,8 @@ class MCPServerService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_servers(cls, tenant_id: str, id_list: list[str] | None, page_number, items_per_page, orderby, desc, keywords):
+    def get_servers(cls, tenant_id: str, id_list: list[str] | None, page_number, items_per_page, orderby, desc,
+                    keywords):
         """Retrieve all MCP servers associated with a tenant.
 
         This method fetches all MCP servers for a given tenant, ordered by creation time.
@@ -84,3 +85,8 @@ class MCPServerService(CommonService):
             return bool(mcp_server), mcp_server
         except Exception:
             return False, None
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id: str):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()

api/db/services/pipeline_operation_log_service.py

@@ -0,0 +1,263 @@
+#
+#  Copyright 2025 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import json
+import logging
+import os
+from datetime import datetime, timedelta
+
+from peewee import fn
+
+from api.db import VALID_PIPELINE_TASK_TYPES, PipelineTaskType
+from api.db.db_models import DB, Document, PipelineOperationLog
+from api.db.services.canvas_service import UserCanvasService
+from api.db.services.common_service import CommonService
+from api.db.services.document_service import DocumentService
+from api.db.services.knowledgebase_service import KnowledgebaseService
+from api.db.services.task_service import GRAPH_RAPTOR_FAKE_DOC_ID
+from api.utils import current_timestamp, datetime_format, get_uuid
+
+
+class PipelineOperationLogService(CommonService):
+    model = PipelineOperationLog
+
+    @classmethod
+    def get_file_logs_fields(cls):
+        return [
+            cls.model.id,
+            cls.model.document_id,
+            cls.model.tenant_id,
+            cls.model.kb_id,
+            cls.model.pipeline_id,
+            cls.model.pipeline_title,
+            cls.model.parser_id,
+            cls.model.document_name,
+            cls.model.document_suffix,
+            cls.model.document_type,
+            cls.model.source_from,
+            cls.model.progress,
+            cls.model.progress_msg,
+            cls.model.process_begin_at,
+            cls.model.process_duration,
+            cls.model.dsl,
+            cls.model.task_type,
+            cls.model.operation_status,
+            cls.model.avatar,
+            cls.model.status,
+            cls.model.create_time,
+            cls.model.create_date,
+            cls.model.update_time,
+            cls.model.update_date,
+        ]
+
+    @classmethod
+    def get_dataset_logs_fields(cls):
+        return [
+            cls.model.id,
+            cls.model.tenant_id,
+            cls.model.kb_id,
+            cls.model.progress,
+            cls.model.progress_msg,
+            cls.model.process_begin_at,
+            cls.model.process_duration,
+            cls.model.task_type,
+            cls.model.operation_status,
+            cls.model.avatar,
+            cls.model.status,
+            cls.model.create_time,
+            cls.model.create_date,
+            cls.model.update_time,
+            cls.model.update_date,
+        ]
+
+    @classmethod
+    def save(cls, **kwargs):
+        """
+        wrap this function in a transaction
+        """
+        sample_obj = cls.model(**kwargs).save(force_insert=True)
+        return sample_obj
+
+    @classmethod
+    @DB.connection_context()
+    def create(cls, document_id, pipeline_id, task_type, fake_document_ids=[], dsl: str = "{}"):
+        referred_document_id = document_id
+
+        if referred_document_id == GRAPH_RAPTOR_FAKE_DOC_ID and fake_document_ids:
+            referred_document_id = fake_document_ids[0]
+        ok, document = DocumentService.get_by_id(referred_document_id)
+        if not ok:
+            logging.warning(f"Document for referred_document_id {referred_document_id} not found")
+            return
+        DocumentService.update_progress_immediately([document.to_dict()])
+        ok, document = DocumentService.get_by_id(referred_document_id)
+        if not ok:
+            logging.warning(f"Document for referred_document_id {referred_document_id} not found")
+            return
+        if document.progress not in [1, -1]:
+            return
+        operation_status = document.run
+
+        if pipeline_id:
+            ok, user_pipeline = UserCanvasService.get_by_id(pipeline_id)
+            if not ok:
+                raise RuntimeError(f"Pipeline {pipeline_id} not found")
+            tenant_id = user_pipeline.user_id
+            title = user_pipeline.title
+            avatar = user_pipeline.avatar
+        else:
+            ok, kb_info = KnowledgebaseService.get_by_id(document.kb_id)
+            if not ok:
+                raise RuntimeError(f"Cannot find knowledge base {document.kb_id} for referred_document {referred_document_id}")
+
+            tenant_id = kb_info.tenant_id
+            title = document.parser_id
+            avatar = document.thumbnail
+
+        if task_type not in VALID_PIPELINE_TASK_TYPES:
+            raise ValueError(f"Invalid task type: {task_type}")
+
+        if task_type in [PipelineTaskType.GRAPH_RAG, PipelineTaskType.RAPTOR, PipelineTaskType.MINDMAP]:
+            finish_at = document.process_begin_at + timedelta(seconds=document.process_duration)
+            if task_type == PipelineTaskType.GRAPH_RAG:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"graphrag_task_finish_at": finish_at},
+                )
+            elif task_type == PipelineTaskType.RAPTOR:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"raptor_task_finish_at": finish_at},
+                )
+            elif task_type == PipelineTaskType.MINDMAP:
+                KnowledgebaseService.update_by_id(
+                    document.kb_id,
+                    {"mindmap_task_finish_at": finish_at},
+                )
+
+        log = dict(
+            id=get_uuid(),
+            document_id=document_id,  # GRAPH_RAPTOR_FAKE_DOC_ID or real document_id
+            tenant_id=tenant_id,
+            kb_id=document.kb_id,
+            pipeline_id=pipeline_id,
+            pipeline_title=title,
+            parser_id=document.parser_id,
+            document_name=document.name,
+            document_suffix=document.suffix,
+            document_type=document.type,
+            source_from="",  # TODO: add in the future
+            progress=document.progress,
+            progress_msg=document.progress_msg,
+            process_begin_at=document.process_begin_at,
+            process_duration=document.process_duration,
+            dsl=json.loads(dsl),
+            task_type=task_type,
+            operation_status=operation_status,
+            avatar=avatar,
+        )
+        log["create_time"] = current_timestamp()
+        log["create_date"] = datetime_format(datetime.now())
+        log["update_time"] = current_timestamp()
+        log["update_date"] = datetime_format(datetime.now())
+
+        with DB.atomic():
+            obj = cls.save(**log)
+
+            limit = int(os.getenv("PIPELINE_OPERATION_LOG_LIMIT", 1000))
+            total = cls.model.select().where(cls.model.kb_id == document.kb_id).count()
+
+            if total > limit:
+                keep_ids = [m.id for m in cls.model.select(cls.model.id).where(cls.model.kb_id == document.kb_id).order_by(cls.model.create_time.desc()).limit(limit)]
+
+                deleted = cls.model.delete().where(cls.model.kb_id == document.kb_id, cls.model.id.not_in(keep_ids)).execute()
+                logging.info(f"[PipelineOperationLogService] Cleaned {deleted} old logs, kept latest {limit} for {document.kb_id}")
+
+        return obj
+
+    @classmethod
+    @DB.connection_context()
+    def record_pipeline_operation(cls, document_id, pipeline_id, task_type, fake_document_ids=[]):
+        return cls.create(document_id=document_id, pipeline_id=pipeline_id, task_type=task_type, fake_document_ids=fake_document_ids)
+
+    @classmethod
+    @DB.connection_context()
+    def get_file_logs_by_kb_id(cls, kb_id, page_number, items_per_page, orderby, desc, keywords, operation_status, types, suffix, create_date_from=None, create_date_to=None):
+        fields = cls.get_file_logs_fields()
+        if keywords:
+            logs = cls.model.select(*fields).where((cls.model.kb_id == kb_id), (fn.LOWER(cls.model.document_name).contains(keywords.lower())))
+        else:
+            logs = cls.model.select(*fields).where(cls.model.kb_id == kb_id)
+
+        logs = logs.where(cls.model.document_id != GRAPH_RAPTOR_FAKE_DOC_ID)
+
+        if operation_status:
+            logs = logs.where(cls.model.operation_status.in_(operation_status))
+        if types:
+            logs = logs.where(cls.model.document_type.in_(types))
+        if suffix:
+            logs = logs.where(cls.model.document_suffix.in_(suffix))
+        if create_date_from:
+            logs = logs.where(cls.model.create_date >= create_date_from)
+        if create_date_to:
+            logs = logs.where(cls.model.create_date <= create_date_to)
+
+        count = logs.count()
+        if desc:
+            logs = logs.order_by(cls.model.getter_by(orderby).desc())
+        else:
+            logs = logs.order_by(cls.model.getter_by(orderby).asc())
+
+        if page_number and items_per_page:
+            logs = logs.paginate(page_number, items_per_page)
+
+        return list(logs.dicts()), count
+
+    @classmethod
+    @DB.connection_context()
+    def get_documents_info(cls, id):
+        fields = [Document.id, Document.name, Document.progress, Document.kb_id]
+        return (
+            cls.model.select(*fields)
+            .join(Document, on=(cls.model.document_id == Document.id))
+            .where(
+                cls.model.id == id
+            )
+            .dicts()
+        )
+
+    @classmethod
+    @DB.connection_context()
+    def get_dataset_logs_by_kb_id(cls, kb_id, page_number, items_per_page, orderby, desc, operation_status, create_date_from=None, create_date_to=None):
+        fields = cls.get_dataset_logs_fields()
+        logs = cls.model.select(*fields).where((cls.model.kb_id == kb_id), (cls.model.document_id == GRAPH_RAPTOR_FAKE_DOC_ID))
+
+        if operation_status:
+            logs = logs.where(cls.model.operation_status.in_(operation_status))
+        if create_date_from:
+            logs = logs.where(cls.model.create_date >= create_date_from)
+        if create_date_to:
+            logs = logs.where(cls.model.create_date <= create_date_to)
+
+        count = logs.count()
+        if desc:
+            logs = logs.order_by(cls.model.getter_by(orderby).desc())
+        else:
+            logs = logs.order_by(cls.model.getter_by(orderby).asc())
+
+        if page_number and items_per_page:
+            logs = logs.paginate(page_number, items_per_page)
+
+        return list(logs.dicts()), count

api/db/services/search_service.py

@@ -94,7 +94,8 @@ class SearchService(CommonService):
         query = (
             cls.model.select(*fields)
             .join(User, on=(cls.model.tenant_id == User.id))
-            .where(((cls.model.tenant_id.in_(joined_tenant_ids)) | (cls.model.tenant_id == user_id)) & (cls.model.status == StatusEnum.VALID.value))
+            .where(((cls.model.tenant_id.in_(joined_tenant_ids)) | (cls.model.tenant_id == user_id)) & (
+                        cls.model.status == StatusEnum.VALID.value))
         )
 
         if keywords:
@@ -110,3 +111,8 @@ class SearchService(CommonService):
             query = query.paginate(page_number, items_per_page)
 
         return list(query.dicts()), count
+
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()

api/db/services/task_service.py

@@ -35,6 +35,8 @@ from rag.utils.redis_conn import REDIS_CONN
 from api import settings
 from rag.nlp import search
 
+CANVAS_DEBUG_DOC_ID = "dataflow_x"
+GRAPH_RAPTOR_FAKE_DOC_ID = "graph_raptor_x"
 
 def trim_header_by_lines(text: str, max_length) -> str:
     # Trim header text to maximum length while preserving line breaks
@@ -70,7 +72,7 @@ class TaskService(CommonService):
 
     @classmethod
     @DB.connection_context()
-    def get_task(cls, task_id):
+    def get_task(cls, task_id, doc_ids=[]):
         """Retrieve detailed task information by task ID.
 
         This method fetches comprehensive task details including associated document,
@@ -84,6 +86,10 @@ class TaskService(CommonService):
             dict: Task details dictionary containing all task information and related metadata.
                  Returns None if task is not found or has exceeded retry limit.
         """
+        doc_id = cls.model.doc_id
+        if doc_id == CANVAS_DEBUG_DOC_ID and doc_ids:
+            doc_id = doc_ids[0]
+
         fields = [
             cls.model.id,
             cls.model.doc_id,
@@ -109,7 +115,7 @@ class TaskService(CommonService):
         ]
         docs = (
             cls.model.select(*fields)
-                .join(Document, on=(cls.model.doc_id == Document.id))
+                .join(Document, on=(doc_id == Document.id))
                 .join(Knowledgebase, on=(Document.kb_id == Knowledgebase.id))
                 .join(Tenant, on=(Knowledgebase.tenant_id == Tenant.id))
                 .where(cls.model.id == task_id)
@@ -159,7 +165,7 @@ class TaskService(CommonService):
         ]
         tasks = (
             cls.model.select(*fields).order_by(cls.model.from_page.asc(), cls.model.create_time.desc())
-                .where(cls.model.doc_id == doc_id)
+            .where(cls.model.doc_id == doc_id)
         )
         tasks = list(tasks.dicts())
         if not tasks:
@@ -199,18 +205,18 @@ class TaskService(CommonService):
                 cls.model.select(
                     *[Document.id, Document.kb_id, Document.location, File.parent_id]
                 )
-                    .join(Document, on=(cls.model.doc_id == Document.id))
+                .join(Document, on=(cls.model.doc_id == Document.id))
-                    .join(
+                .join(
                     File2Document,
                     on=(File2Document.document_id == Document.id),
                     join_type=JOIN.LEFT_OUTER,
                 )
-                    .join(
+                .join(
                     File,
                     on=(File2Document.file_id == File.id),
                     join_type=JOIN.LEFT_OUTER,
                 )
-                    .where(
+                .where(
                     Document.status == StatusEnum.VALID.value,
                     Document.run == TaskStatus.RUNNING.value,
                     ~(Document.type == FileType.VIRTUAL.value),
@@ -288,25 +294,33 @@ class TaskService(CommonService):
                 cls.model.update(progress=prog).where(
                     (cls.model.id == id) &
                     (
-                        (cls.model.progress != -1) &
+                            (cls.model.progress != -1) &
-                        ((prog == -1) | (prog > cls.model.progress))
+                            ((prog == -1) | (prog > cls.model.progress))
                     )
                 ).execute()
-            return
+        else:
+            with DB.lock("update_progress", -1):
+                if info["progress_msg"]:
+                    progress_msg = trim_header_by_lines(task.progress_msg + "\n" + info["progress_msg"], 3000)
+                    cls.model.update(progress_msg=progress_msg).where(cls.model.id == id).execute()
+                if "progress" in info:
+                    prog = info["progress"]
+                    cls.model.update(progress=prog).where(
+                        (cls.model.id == id) &
+                        (
+                            (cls.model.progress != -1) &
+                            ((prog == -1) | (prog > cls.model.progress))
+                        )
+                    ).execute()
 
-        with DB.lock("update_progress", -1):
+        process_duration = (datetime.now() - task.begin_at).total_seconds()
-            if info["progress_msg"]:
+        cls.model.update(process_duration=process_duration).where(cls.model.id == id).execute()
-                progress_msg = trim_header_by_lines(task.progress_msg + "\n" + info["progress_msg"], 3000)
+
-                cls.model.update(progress_msg=progress_msg).where(cls.model.id == id).execute()
+    @classmethod
-            if "progress" in info:
+    @DB.connection_context()
-                prog = info["progress"]
+    def delete_by_doc_ids(cls, doc_ids):
-                cls.model.update(progress=prog).where(
+        """Delete task associated with a document."""
-                    (cls.model.id == id) &
+        return cls.model.delete().where(cls.model.doc_id.in_(doc_ids)).execute()
-                    (
-                        (cls.model.progress != -1) &
-                        ((prog == -1) | (prog > cls.model.progress))
-                    )
-                ).execute()
 
 
 def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
@@ -329,8 +343,16 @@ def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
         - Task digests are calculated for optimization and reuse
         - Previous task chunks may be reused if available
     """
+
     def new_task():
-        return {"id": get_uuid(), "doc_id": doc["id"], "progress": 0.0, "from_page": 0, "to_page": 100000000}
+        return {
+            "id": get_uuid(),
+            "doc_id": doc["id"],
+            "progress": 0.0,
+            "from_page": 0,
+            "to_page": 100000000,
+            "begin_at": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+        }
 
     parse_task_array = []
 
@@ -343,7 +365,7 @@ def queue_tasks(doc: dict, bucket: str, name: str, priority: int):
         page_size = doc["parser_config"].get("task_page_size") or 12
         if doc["parser_id"] == "paper":
             page_size = doc["parser_config"].get("task_page_size") or 22
-        if doc["parser_id"] in ["one", "knowledge_graph"] or do_layout != "DeepDOC":
+        if doc["parser_id"] in ["one", "knowledge_graph"] or do_layout != "DeepDOC" or doc["parser_config"].get("toc", True):
             page_size = 10 ** 9
         page_ranges = doc["parser_config"].get("pages") or [(1, 10 ** 5)]
         for s, e in page_ranges:
@@ -472,36 +494,29 @@ def has_canceled(task_id):
     return False
 
 
-def queue_dataflow(dsl:str, tenant_id:str, doc_id:str, task_id:str, flow_id:str, priority: int, callback=None) -> tuple[bool, str]:
+def queue_dataflow(tenant_id:str, flow_id:str, task_id:str, doc_id:str=CANVAS_DEBUG_DOC_ID, file:dict=None, priority: int=0, rerun:bool=False) -> tuple[bool, str]:
-    """
-    Returns a tuple (success: bool, error_message: str).
-    """
-    _ = callback
 
     task = dict(
-    id=get_uuid() if not task_id else task_id,
+        id=task_id,
-    doc_id=doc_id,
+        doc_id=doc_id,
-    from_page=0,
+        from_page=0,
-    to_page=100000000,
+        to_page=100000000,
-    task_type="dataflow",
+        task_type="dataflow" if not rerun else "dataflow_rerun",
-    priority=priority,
+        priority=priority,
+        begin_at= datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
     )
-
+    if doc_id not in [CANVAS_DEBUG_DOC_ID, GRAPH_RAPTOR_FAKE_DOC_ID]:
-    TaskService.model.delete().where(TaskService.model.id == task["id"]).execute()
+        TaskService.model.delete().where(TaskService.model.doc_id == doc_id).execute()
+        DocumentService.begin2parse(doc_id)
     bulk_insert_into_db(model=Task, data_source=[task], replace_on_conflict=True)
 
-    kb_id = DocumentService.get_knowledgebase_id(doc_id)
+    task["kb_id"] = DocumentService.get_knowledgebase_id(doc_id)
-    if not kb_id:
-        return False, f"Can't find KB of this document: {doc_id}"
-
-    task["kb_id"] = kb_id
     task["tenant_id"] = tenant_id
-    task["task_type"] = "dataflow"
+    task["dataflow_id"] = flow_id
-    task["dsl"] = dsl
+    task["file"] = file
-    task["dataflow_id"] = get_uuid() if not flow_id else flow_id
 
     if not REDIS_CONN.queue_product(
-        get_svr_queue_name(priority), message=task
+            get_svr_queue_name(priority), message=task
     ):
         return False, "Can't access Redis. Please check the Redis' status."
 

api/db/services/tenant_llm_service.py

@@ -57,8 +57,10 @@ class TenantLLMService(CommonService):
     @classmethod
     @DB.connection_context()
     def get_my_llms(cls, tenant_id):
-        fields = [cls.model.llm_factory, LLMFactories.logo, LLMFactories.tags, cls.model.model_type, cls.model.llm_name, cls.model.used_tokens]
+        fields = [cls.model.llm_factory, LLMFactories.logo, LLMFactories.tags, cls.model.model_type, cls.model.llm_name,
-        objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory == LLMFactories.name)).where(cls.model.tenant_id == tenant_id, ~cls.model.api_key.is_null()).dicts()
+                  cls.model.used_tokens]
+        objs = cls.model.select(*fields).join(LLMFactories, on=(cls.model.llm_factory == LLMFactories.name)).where(
+            cls.model.tenant_id == tenant_id, ~cls.model.api_key.is_null()).dicts()
 
         return list(objs)
 
@@ -122,7 +124,8 @@ class TenantLLMService(CommonService):
                     model_config = {"llm_factory": llm[0].fid, "api_key": "", "llm_name": mdlnm, "api_base": ""}
             if not model_config:
                 if mdlnm == "flag-embedding":
-                    model_config = {"llm_factory": "Tongyi-Qianwen", "api_key": "", "llm_name": llm_name, "api_base": ""}
+                    model_config = {"llm_factory": "Tongyi-Qianwen", "api_key": "", "llm_name": llm_name,
+                                    "api_base": ""}
                 else:
                     if not mdlnm:
                         raise LookupError(f"Type of {llm_type} model is not set.")
@@ -137,27 +140,33 @@ class TenantLLMService(CommonService):
         if llm_type == LLMType.EMBEDDING.value:
             if model_config["llm_factory"] not in EmbeddingModel:
                 return
-            return EmbeddingModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"], base_url=model_config["api_base"])
+            return EmbeddingModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"],
+                                                               base_url=model_config["api_base"])
 
         if llm_type == LLMType.RERANK:
             if model_config["llm_factory"] not in RerankModel:
                 return
-            return RerankModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"], base_url=model_config["api_base"])
+            return RerankModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"],
+                                                            base_url=model_config["api_base"])
 
         if llm_type == LLMType.IMAGE2TEXT.value:
             if model_config["llm_factory"] not in CvModel:
                 return
-            return CvModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"], lang, base_url=model_config["api_base"], **kwargs)
+            return CvModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"], lang,
+                                                        base_url=model_config["api_base"], **kwargs)
 
         if llm_type == LLMType.CHAT.value:
             if model_config["llm_factory"] not in ChatModel:
                 return
-            return ChatModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"], base_url=model_config["api_base"], **kwargs)
+            return ChatModel[model_config["llm_factory"]](model_config["api_key"], model_config["llm_name"],
+                                                          base_url=model_config["api_base"], **kwargs)
 
         if llm_type == LLMType.SPEECH2TEXT:
             if model_config["llm_factory"] not in Seq2txtModel:
                 return
-            return Seq2txtModel[model_config["llm_factory"]](key=model_config["api_key"], model_name=model_config["llm_name"], lang=lang, base_url=model_config["api_base"])
+            return Seq2txtModel[model_config["llm_factory"]](key=model_config["api_key"],
+                                                             model_name=model_config["llm_name"], lang=lang,
+                                                             base_url=model_config["api_base"])
         if llm_type == LLMType.TTS:
             if model_config["llm_factory"] not in TTSModel:
                 return
@@ -194,11 +203,14 @@ class TenantLLMService(CommonService):
         try:
             num = (
                 cls.model.update(used_tokens=cls.model.used_tokens + used_tokens)
-                .where(cls.model.tenant_id == tenant_id, cls.model.llm_name == llm_name, cls.model.llm_factory == llm_factory if llm_factory else True)
+                .where(cls.model.tenant_id == tenant_id, cls.model.llm_name == llm_name,
+                       cls.model.llm_factory == llm_factory if llm_factory else True)
                 .execute()
             )
         except Exception:
-            logging.exception("TenantLLMService.increase_usage got exception,Failed to update used_tokens for tenant_id=%s, llm_name=%s", tenant_id, llm_name)
+            logging.exception(
+                "TenantLLMService.increase_usage got exception,Failed to update used_tokens for tenant_id=%s, llm_name=%s",
+                tenant_id, llm_name)
             return 0
 
         return num
@@ -206,9 +218,16 @@ class TenantLLMService(CommonService):
     @classmethod
     @DB.connection_context()
     def get_openai_models(cls):
-        objs = cls.model.select().where((cls.model.llm_factory == "OpenAI"), ~(cls.model.llm_name == "text-embedding-3-small"), ~(cls.model.llm_name == "text-embedding-3-large")).dicts()
+        objs = cls.model.select().where((cls.model.llm_factory == "OpenAI"),
+                                        ~(cls.model.llm_name == "text-embedding-3-small"),
+                                        ~(cls.model.llm_name == "text-embedding-3-large")).dicts()
         return list(objs)
 
+    @classmethod
+    @DB.connection_context()
+    def delete_by_tenant_id(cls, tenant_id):
+        return cls.model.delete().where(cls.model.tenant_id == tenant_id).execute()
+
     @staticmethod
     def llm_id2llm_type(llm_id: str) -> str | None:
         from api.db.services.llm_service import LLMService
@@ -245,8 +264,9 @@ class LLM4Tenant:
         langfuse_keys = TenantLangfuseService.filter_by_tenant(tenant_id=tenant_id)
         self.langfuse = None
         if langfuse_keys:
-            langfuse = Langfuse(public_key=langfuse_keys.public_key, secret_key=langfuse_keys.secret_key, host=langfuse_keys.host)
+            langfuse = Langfuse(public_key=langfuse_keys.public_key, secret_key=langfuse_keys.secret_key,
+                                host=langfuse_keys.host)
             if langfuse.auth_check():
                 self.langfuse = langfuse
                 trace_id = self.langfuse.create_trace_id()
-                self.trace_context = {"trace_id": trace_id}
+                self.trace_context = {"trace_id": trace_id}

api/db/services/user_canvas_version.py

@@ -2,45 +2,60 @@ from api.db.db_models import UserCanvasVersion, DB
 from api.db.services.common_service import CommonService
 from peewee import DoesNotExist
 
+
 class UserCanvasVersionService(CommonService):
     model = UserCanvasVersion
-    
+
-    
     @classmethod
     @DB.connection_context()
     def list_by_canvas_id(cls, user_canvas_id):
         try:
             user_canvas_version = cls.model.select(
-                *[cls.model.id, 
+                *[cls.model.id,
-                cls.model.create_time,
+                  cls.model.create_time,
-                cls.model.title,
+                  cls.model.title,
-                cls.model.create_date, 
+                  cls.model.create_date,
-                cls.model.update_date,
+                  cls.model.update_date,
-                cls.model.user_canvas_id, 
+                  cls.model.user_canvas_id,
-                cls.model.update_time]
+                  cls.model.update_time]
             ).where(cls.model.user_canvas_id == user_canvas_id)
             return user_canvas_version
         except DoesNotExist:
             return None
         except Exception:
             return None
-    
+
+    @classmethod
+    @DB.connection_context()
+    def get_all_canvas_version_by_canvas_ids(cls, canvas_ids):
+        fields = [cls.model.id]
+        versions = cls.model.select(*fields).where(cls.model.user_canvas_id.in_(canvas_ids))
+        versions.order_by(cls.model.create_time.asc())
+        offset, limit = 0, 100
+        res = []
+        while True:
+            version_batch = versions.offset(offset).limit(limit)
+            _temp = list(version_batch.dicts())
+            if not _temp:
+                break
+            res.extend(_temp)
+            offset += limit
+        return res
+
     @classmethod
     @DB.connection_context()
     def delete_all_versions(cls, user_canvas_id):
         try:
-            user_canvas_version = cls.model.select().where(cls.model.user_canvas_id == user_canvas_id).order_by(cls.model.create_time.desc())
+            user_canvas_version = cls.model.select().where(cls.model.user_canvas_id == user_canvas_id).order_by(
+                cls.model.create_time.desc())
             if user_canvas_version.count() > 20:
                 delete_ids = []
                 for i in range(20, user_canvas_version.count()):
                     delete_ids.append(user_canvas_version[i].id)
-                
+
                 cls.delete_by_ids(delete_ids)
             return True
         except DoesNotExist:
             return None
         except Exception:
             return None
-
-
-

api/db/services/user_service.py

@@ -288,6 +288,17 @@ class UserTenantService(CommonService):
                     .join(User, on=((cls.model.tenant_id == User.id) & (UserTenant.user_id == user_id) & (UserTenant.status == StatusEnum.VALID.value)))
                     .where(cls.model.status == StatusEnum.VALID.value).dicts())
 
+    @classmethod
+    @DB.connection_context()
+    def get_user_tenant_relation_by_user_id(cls, user_id):
+        fields = [
+            cls.model.id,
+            cls.model.user_id,
+            cls.model.tenant_id,
+            cls.model.role
+        ]
+        return list(cls.model.select(*fields).where(cls.model.user_id == user_id).dicts().dicts())
+
     @classmethod
     @DB.connection_context()
     def get_num_members(cls, user_id: str):
@@ -304,4 +315,4 @@ class UserTenantService(CommonService):
             ).first()
             return user_tenant
         except peewee.DoesNotExist:
-            return None
+            return None

api/settings.py

@@ -65,8 +65,8 @@ OAUTH_CONFIG = None
 DOC_ENGINE = None
 docStoreConn = None
 
-retrievaler = None
+retriever = None
-kg_retrievaler = None
+kg_retriever = None
 
 # user registration switch
 REGISTER_ENABLED = 1
@@ -174,7 +174,7 @@ def init_settings():
 
     OAUTH_CONFIG = get_base_config("oauth", {})
 
-    global DOC_ENGINE, docStoreConn, retrievaler, kg_retrievaler
+    global DOC_ENGINE, docStoreConn, retriever, kg_retriever
     DOC_ENGINE = os.environ.get("DOC_ENGINE", "elasticsearch")
     # DOC_ENGINE = os.environ.get('DOC_ENGINE', "opensearch")
     lower_case_doc_engine = DOC_ENGINE.lower()
@@ -187,10 +187,10 @@ def init_settings():
     else:
         raise Exception(f"Not supported doc engine: {DOC_ENGINE}")
 
-    retrievaler = search.Dealer(docStoreConn)
+    retriever = search.Dealer(docStoreConn)
     from graphrag import search as kg_search
 
-    kg_retrievaler = kg_search.KGSearch(docStoreConn)
+    kg_retriever = kg_search.KGSearch(docStoreConn)
 
     if int(os.environ.get("SANDBOX_ENABLED", "0")):
         global SANDBOX_HOST

api/utils/api_utils.py

@@ -51,15 +51,13 @@ from api import settings
 from api.constants import REQUEST_MAX_WAIT_SEC, REQUEST_WAIT_SEC
 from api.db import ActiveEnum
 from api.db.db_models import APIToken
-from api.db.services import UserService
-from api.db.services.llm_service import LLMService
-from api.db.services.tenant_llm_service import TenantLLMService
 from api.utils.json import CustomJSONEncoder, json_dumps
 from api.utils import get_uuid
 from rag.utils.mcp_tool_call_conn import MCPToolCallSession, close_multiple_mcp_toolcall_sessions
 
 requests.models.complexjson.dumps = functools.partial(json.dumps, cls=CustomJSONEncoder)
 
+
 def serialize_for_json(obj):
     """
     Recursively serialize objects to make them JSON serializable.
@@ -68,8 +66,8 @@ def serialize_for_json(obj):
     if hasattr(obj, '__dict__'):
         # For objects with __dict__, try to serialize their attributes
         try:
-            return {key: serialize_for_json(value) for key, value in obj.__dict__.items() 
+            return {key: serialize_for_json(value) for key, value in obj.__dict__.items()
-                   if not key.startswith('_')}
+                    if not key.startswith('_')}
         except (AttributeError, TypeError):
             return str(obj)
     elif hasattr(obj, '__name__'):
@@ -85,6 +83,7 @@ def serialize_for_json(obj):
         # Fallback: convert to string representation
         return str(obj)
 
+
 def request(**kwargs):
     sess = requests.Session()
     stream = kwargs.pop("stream", sess.stream)
@@ -105,7 +104,8 @@ def request(**kwargs):
                         settings.HTTP_APP_KEY.encode("ascii"),
                         prepped.path_url.encode("ascii"),
                         prepped.body if kwargs.get("json") else b"",
-                        urlencode(sorted(kwargs["data"].items()), quote_via=quote, safe="-._~").encode("ascii") if kwargs.get("data") and isinstance(kwargs["data"], dict) else b"",
+                        urlencode(sorted(kwargs["data"].items()), quote_via=quote, safe="-._~").encode(
+                            "ascii") if kwargs.get("data") and isinstance(kwargs["data"], dict) else b"",
                     ]
                 ),
                 "sha1",
@@ -127,7 +127,7 @@ def request(**kwargs):
 def get_exponential_backoff_interval(retries, full_jitter=False):
     """Calculate the exponential backoff wait time."""
     # Will be zero if factor equals 0
-    countdown = min(REQUEST_MAX_WAIT_SEC, REQUEST_WAIT_SEC * (2**retries))
+    countdown = min(REQUEST_MAX_WAIT_SEC, REQUEST_WAIT_SEC * (2 ** retries))
     # Full jitter according to
     # https://aws.amazon.com/blogs/architecture/exponential-backoff-and-jitter/
     if full_jitter:
@@ -151,18 +151,21 @@ def get_data_error_result(code=settings.RetCode.DATA_ERROR, message="Sorry! Data
 def server_error_response(e):
     logging.exception(e)
     try:
-        if e.code == 401:
+        msg = repr(e).lower()
-            return get_json_result(code=401, message=repr(e))
+        if getattr(e, "code", None) == 401 or ("unauthorized" in msg) or ("401" in msg):
-    except BaseException:
+            return get_json_result(code=settings.RetCode.UNAUTHORIZED, message=repr(e))
-        pass
+    except Exception as ex:
+        logging.warning(f"error checking authorization: {ex}")
+
     if len(e.args) > 1:
         try:
             serialized_data = serialize_for_json(e.args[1])
-            return get_json_result(code= settings.RetCode.EXCEPTION_ERROR, message=repr(e.args[0]), data=serialized_data)
+            return get_json_result(code=settings.RetCode.EXCEPTION_ERROR, message=repr(e.args[0]), data=serialized_data)
         except Exception:
             return get_json_result(code=settings.RetCode.EXCEPTION_ERROR, message=repr(e.args[0]), data=None)
     if repr(e).find("index_not_found_exception") >= 0:
-        return get_json_result(code=settings.RetCode.EXCEPTION_ERROR, message="No chunk found, please upload file and parse it.")
+        return get_json_result(code=settings.RetCode.EXCEPTION_ERROR,
+                               message="No chunk found, please upload file and parse it.")
 
     return get_json_result(code=settings.RetCode.EXCEPTION_ERROR, message=repr(e))
 
@@ -207,7 +210,8 @@ def validate_request(*args, **kwargs):
                 if no_arguments:
                     error_string += "required argument are missing: {}; ".format(",".join(no_arguments))
                 if error_arguments:
-                    error_string += "required argument values: {}".format(",".join(["{}={}".format(a[0], a[1]) for a in error_arguments]))
+                    error_string += "required argument values: {}".format(
+                        ",".join(["{}={}".format(a[0], a[1]) for a in error_arguments]))
                 return get_json_result(code=settings.RetCode.ARGUMENT_ERROR, message=error_string)
             return func(*_args, **_kwargs)
 
@@ -222,7 +226,8 @@ def not_allowed_parameters(*params):
             input_arguments = flask_request.json or flask_request.form.to_dict()
             for param in params:
                 if param in input_arguments:
-                    return get_json_result(code=settings.RetCode.ARGUMENT_ERROR, message=f"Parameter {param} isn't allowed")
+                    return get_json_result(code=settings.RetCode.ARGUMENT_ERROR,
+                                           message=f"Parameter {param} isn't allowed")
             return f(*args, **kwargs)
 
         return wrapper
@@ -233,12 +238,14 @@ def not_allowed_parameters(*params):
 def active_required(f):
     @wraps(f)
     def wrapper(*args, **kwargs):
+        from api.db.services import UserService
         user_id = current_user.id
         usr = UserService.filter_by_id(user_id)
         # check is_active
         if not usr or not usr.is_active == ActiveEnum.ACTIVE.value:
             return get_json_result(code=settings.RetCode.FORBIDDEN, message="User isn't active, please activate first.")
         return f(*args, **kwargs)
+
     return wrapper
 
 
@@ -259,7 +266,7 @@ def send_file_in_mem(data, filename):
     return send_file(f, as_attachment=True, attachment_filename=filename)
 
 
-def get_json_result(code=settings.RetCode.SUCCESS, message="success", data=None):
+def get_json_result(code: settings.RetCode = settings.RetCode.SUCCESS, message="success", data=None):
     response = {"code": code, "message": message, "data": data}
     return jsonify(response)
 
@@ -314,7 +321,7 @@ def construct_result(code=settings.RetCode.DATA_ERROR, message="data is missing"
     return jsonify(response)
 
 
-def construct_json_result(code=settings.RetCode.SUCCESS, message="success", data=None):
+def construct_json_result(code: settings.RetCode = settings.RetCode.SUCCESS, message="success", data=None):
     if data is None:
         return jsonify({"code": code, "message": message})
     else:
@@ -347,27 +354,39 @@ def token_required(func):
         token = authorization_list[1]
         objs = APIToken.query(token=token)
         if not objs:
-            return get_json_result(data=False, message="Authentication error: API key is invalid!", code=settings.RetCode.AUTHENTICATION_ERROR)
+            return get_json_result(data=False, message="Authentication error: API key is invalid!",
+                                   code=settings.RetCode.AUTHENTICATION_ERROR)
         kwargs["tenant_id"] = objs[0].tenant_id
         return func(*args, **kwargs)
 
     return decorated_function
 
 
-def get_result(code=settings.RetCode.SUCCESS, message="", data=None):
+def get_result(code=settings.RetCode.SUCCESS, message="", data=None, total=None):
-    if code == 0:
+    """
+    Standard API response format:
+    {
+        "code": 0,
+        "data": [...],        # List or object, backward compatible
+        "total": 47,          # Optional field for pagination
+        "message": "..."      # Error or status message
+    }
+    """
+    response = {"code": code}
+
+    if code == settings.RetCode.SUCCESS:
         if data is not None:
-            response = {"code": code, "data": data}
+            response["data"] = data
-        else:
+        if total is not None:
-            response = {"code": code}
+            response["total_datasets"] = total
     else:
-        response = {"code": code, "message": message}
+        response["message"] = message or "Error"
+
     return jsonify(response)
 
-
 def get_error_data_result(
-    message="Sorry! Data missing!",
+        message="Sorry! Data missing!",
-    code=settings.RetCode.DATA_ERROR,
+        code=settings.RetCode.DATA_ERROR,
 ):
     result_dict = {"code": code, "message": message}
     response = {}
@@ -402,7 +421,8 @@ def get_parser_config(chunk_method, parser_config):
 
     # Define default configurations for each chunking method
     key_mapping = {
-        "naive": {"chunk_token_num": 512, "delimiter": r"\n", "html4excel": False, "layout_recognize": "DeepDOC", "raptor": {"use_raptor": False}, "graphrag": {"use_graphrag": False}},
+        "naive": {"chunk_token_num": 512, "delimiter": r"\n", "html4excel": False, "layout_recognize": "DeepDOC",
+                  "raptor": {"use_raptor": False}, "graphrag": {"use_graphrag": False}},
         "qa": {"raptor": {"use_raptor": False}, "graphrag": {"use_graphrag": False}},
         "tag": None,
         "resume": None,
@@ -441,16 +461,16 @@ def get_parser_config(chunk_method, parser_config):
 
 
 def get_data_openai(
-    id=None,
+        id=None,
-    created=None,
+        created=None,
-    model=None,
+        model=None,
-    prompt_tokens=0,
+        prompt_tokens=0,
-    completion_tokens=0,
+        completion_tokens=0,
-    content=None,
+        content=None,
-    finish_reason=None,
+        finish_reason=None,
-    object="chat.completion",
+        object="chat.completion",
-    param=None,
+        param=None,
-    stream=False
+        stream=False
 ):
     total_tokens = prompt_tokens + completion_tokens
 
@@ -524,6 +544,8 @@ def check_duplicate_ids(ids, id_type="item"):
 
 
 def verify_embedding_availability(embd_id: str, tenant_id: str) -> tuple[bool, Response | None]:
+    from api.db.services.llm_service import LLMService
+    from api.db.services.tenant_llm_service import TenantLLMService
     """
     Verifies availability of an embedding model for a specific tenant.
 
@@ -562,7 +584,9 @@ def verify_embedding_availability(embd_id: str, tenant_id: str) -> tuple[bool, R
         in_llm_service = bool(LLMService.query(llm_name=llm_name, fid=llm_factory, model_type="embedding"))
 
         tenant_llms = TenantLLMService.get_my_llms(tenant_id=tenant_id)
-        is_tenant_model = any(llm["llm_name"] == llm_name and llm["llm_factory"] == llm_factory and llm["model_type"] == "embedding" for llm in tenant_llms)
+        is_tenant_model = any(
+            llm["llm_name"] == llm_name and llm["llm_factory"] == llm_factory and llm["model_type"] == "embedding" for
+            llm in tenant_llms)
 
         is_builtin_model = embd_id in settings.BUILTIN_EMBEDDING_MODELS
         if not (is_builtin_model or is_tenant_model or in_llm_service):
@@ -659,6 +683,16 @@ def remap_dictionary_keys(source_data: dict, key_aliases: dict = None) -> dict:
     return transformed_data
 
 
+def group_by(list_of_dict, key):
+    res = {}
+    for item in list_of_dict:
+        if item[key] in res.keys():
+            res[item[key]].append(item)
+        else:
+            res[item[key]] = [item]
+    return res
+
+
 def get_mcp_tools(mcp_servers: list, timeout: float | int = 10) -> tuple[dict, str]:
     results = {}
     tool_call_sessions = []
@@ -695,7 +729,9 @@ TimeoutException = Union[Type[BaseException], BaseException]
 OnTimeoutCallback = Union[Callable[..., Any], Coroutine[Any, Any, Any]]
 
 
-def timeout(seconds: float | int = None, attempts: int = 2, *, exception: Optional[TimeoutException] = None, on_timeout: Optional[OnTimeoutCallback] = None):
+def timeout(seconds: float | int | str = None, attempts: int = 2, *, exception: Optional[TimeoutException] = None, on_timeout: Optional[OnTimeoutCallback] = None):
+    if isinstance(seconds, str):
+        seconds = float(seconds)
     def decorator(func):
         @wraps(func)
         def wrapper(*args, **kwargs):
@@ -781,7 +817,9 @@ async def is_strong_enough(chat_model, embedding_model):
                 _ = await trio.to_thread.run_sync(lambda: embedding_model.encode(["Are you strong enough!?"]))
         if chat_model:
             with trio.fail_after(30):
-                res = await trio.to_thread.run_sync(lambda: chat_model.chat("Nothing special.", [{"role": "user", "content": "Are you strong enough!?"}], {}))
+                res = await trio.to_thread.run_sync(lambda: chat_model.chat("Nothing special.", [{"role": "user",
+                                                                                                  "content": "Are you strong enough!?"}],
+                                                                            {}))
             if res.find("**ERROR**") >= 0:
                 raise Exception(res)
 

api/utils/base64_image.py

@@ -1,3 +1,56 @@
 import base64
+import logging
+from functools import partial
+from io import BytesIO
+
+from PIL import Image
+
 test_image_base64 = "iVBORw0KGgoAAAANSUhEUgAAAGQAAABkCAIAAAD/gAIDAAAA6ElEQVR4nO3QwQ3AIBDAsIP9d25XIC+EZE8QZc18w5l9O+AlZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBWYFZgVmBT+IYAHHLHkdEgAAAABJRU5ErkJggg=="
-test_image = base64.b64decode(test_image_base64)
+test_image = base64.b64decode(test_image_base64)
+
+
+async def image2id(d: dict, storage_put_func: partial, objname:str, bucket:str="imagetemps"):
+    import logging
+    from io import BytesIO
+    import trio
+    from rag.svr.task_executor import minio_limiter
+    if not d.get("image"):
+        return
+
+    with BytesIO() as output_buffer:
+        if isinstance(d["image"], bytes):
+            output_buffer.write(d["image"])
+            output_buffer.seek(0)
+        else:
+            # If the image is in RGBA mode, convert it to RGB mode before saving it in JPEG format.
+            if d["image"].mode in ("RGBA", "P"):
+                converted_image = d["image"].convert("RGB")
+                d["image"] = converted_image
+            try:
+                d["image"].save(output_buffer, format='JPEG')
+            except OSError as e:
+                logging.warning(
+                    "Saving image exception, ignore: {}".format(str(e)))
+
+        async with minio_limiter:
+            await trio.to_thread.run_sync(lambda: storage_put_func(bucket=bucket, fnm=objname, binary=output_buffer.getvalue()))
+        d["img_id"] = f"{bucket}-{objname}"
+        if not isinstance(d["image"], bytes):
+            d["image"].close()
+        del d["image"]  # Remove image reference
+
+
+def id2image(image_id:str|None, storage_get_func: partial):
+    if not image_id:
+        return
+    arr = image_id.split("-")
+    if len(arr) != 2:
+        return
+    bkt, nm = image_id.split("-")
+    try:
+        blob = storage_get_func(bucket=bkt, filename=nm)
+        if not blob:
+            return
+        return Image.open(BytesIO(blob))
+    except Exception as e:
+        logging.exception(e)

api/utils/common.py

@@ -21,3 +21,26 @@ def string_to_bytes(string):
 
 def bytes_to_string(byte):
     return byte.decode(encoding="utf-8")
+
+
+def convert_bytes(size_in_bytes: int) -> str:
+    """
+    Format size in bytes.
+    """
+    if size_in_bytes == 0:
+        return "0 B"
+
+    units = ['B', 'KB', 'MB', 'GB', 'TB', 'PB']
+    i = 0
+    size = float(size_in_bytes)
+
+    while size >= 1024 and i < len(units) - 1:
+        size /= 1024
+        i += 1
+
+    if i == 0 or size >= 100:
+        return f"{size:.0f} {units[i]}"
+    elif size >= 10:
+        return f"{size:.1f} {units[i]}"
+    else:
+        return f"{size:.2f} {units[i]}"

api/utils/email_templates.py

@@ -0,0 +1,25 @@
+"""
+Reusable HTML email templates and registry.
+"""
+
+# Invitation email template
+INVITE_EMAIL_TMPL = """
+<p>Hi {{email}},</p>
+<p>{{inviter}} has invited you to join their team (ID: {{tenant_id}}).</p>
+<p>Click the link below to complete your registration:<br>
+<a href="{{invite_url}}">{{invite_url}}</a></p>
+<p>If you did not request this, please ignore this email.</p>
+"""
+
+# Password reset code template
+RESET_CODE_EMAIL_TMPL = """
+<p>Hello,</p>
+<p>Your password reset code is: <b>{{ code }}</b></p>
+<p>This code will expire in {{ ttl_min }} minutes.</p>
+"""
+
+# Template registry
+EMAIL_TEMPLATES = {
+    "invite": INVITE_EMAIL_TMPL,
+    "reset_code": RESET_CODE_EMAIL_TMPL,
+}

api/utils/file_utils.py

@@ -13,7 +13,12 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+
+
+# Standard library imports
 import base64
+import hashlib
+import io
 import json
 import os
 import re
@@ -22,13 +27,20 @@ import subprocess
 import sys
 import tempfile
 import threading
+import zipfile
 from io import BytesIO
 
+# Typing
+from typing import List, Union, Tuple
+
+# Third-party imports
+import olefile
 import pdfplumber
 from cachetools import LRUCache, cached
 from PIL import Image
 from ruamel.yaml import YAML
 
+# Local imports
 from api.constants import IMG_BASE64_PREFIX
 from api.db import FileType
 
@@ -155,13 +167,13 @@ def filename_type(filename):
     if re.match(r".*\.pdf$", filename):
         return FileType.PDF.value
 
-    if re.match(r".*\.(eml|doc|docx|ppt|pptx|yml|xml|htm|json|jsonl|ldjson|csv|txt|ini|xls|xlsx|wps|rtf|hlp|pages|numbers|key|md|py|js|java|c|cpp|h|php|go|ts|sh|cs|kt|html|sql)$", filename):
+    if re.match(r".*\.(msg|eml|doc|docx|ppt|pptx|yml|xml|htm|json|jsonl|ldjson|csv|txt|ini|xls|xlsx|wps|rtf|hlp|pages|numbers|key|md|py|js|java|c|cpp|h|php|go|ts|sh|cs|kt|html|sql)$", filename):
         return FileType.DOC.value
 
     if re.match(r".*\.(wav|flac|ape|alac|wavpack|wv|mp3|aac|ogg|vorbis|opus)$", filename):
         return FileType.AURAL.value
 
-    if re.match(r".*\.(jpg|jpeg|png|tif|gif|pcx|tga|exif|fpx|svg|psd|cdr|pcd|dxf|ufo|eps|ai|raw|WMF|webp|avif|apng|icon|ico|mpg|mpeg|avi|rm|rmvb|mov|wmv|asf|dat|asx|wvx|mpe|mpa|mp4)$", filename):
+    if re.match(r".*\.(jpg|jpeg|png|tif|gif|pcx|tga|exif|fpx|svg|psd|cdr|pcd|dxf|ufo|eps|ai|raw|WMF|webp|avif|apng|icon|ico|mpg|mpeg|avi|rm|rmvb|mov|wmv|asf|dat|asx|wvx|mpe|mpa|mp4|avi|mkv)$", filename):
         return FileType.VISUAL.value
 
     return FileType.OTHER.value
@@ -284,3 +296,125 @@ def read_potential_broken_pdf(blob):
         return repaired
 
     return blob
+
+
+
+def _is_zip(h: bytes) -> bool:
+    return h.startswith(b"PK\x03\x04") or h.startswith(b"PK\x05\x06") or h.startswith(b"PK\x07\x08")
+
+def _is_pdf(h: bytes) -> bool:
+    return h.startswith(b"%PDF-")
+
+def _is_ole(h: bytes) -> bool:
+    return h.startswith(b"\xD0\xCF\x11\xE0\xA1\xB1\x1A\xE1")
+
+def _sha10(b: bytes) -> str:
+    return hashlib.sha256(b).hexdigest()[:10]
+
+def _guess_ext(b: bytes) -> str:
+    h = b[:8]
+    if _is_zip(h):
+        try:
+            with zipfile.ZipFile(io.BytesIO(b), "r") as z:
+                names = [n.lower() for n in z.namelist()]
+                if any(n.startswith("word/") for n in names):
+                    return ".docx"
+                if any(n.startswith("ppt/") for n in names):
+                    return ".pptx"
+                if any(n.startswith("xl/") for n in names):
+                    return ".xlsx"
+        except Exception:
+            pass
+        return ".zip"
+    if _is_pdf(h):
+        return ".pdf"
+    if _is_ole(h):
+        return ".doc"
+    return ".bin"
+
+# Try to extract the real embedded payload from OLE's Ole10Native
+def _extract_ole10native_payload(data: bytes) -> bytes:
+    try:
+        pos = 0
+        if len(data) < 4:
+            return data
+        _ = int.from_bytes(data[pos:pos+4], "little")
+        pos += 4
+        # filename/src/tmp (NUL-terminated ANSI)
+        for _ in range(3):
+            z = data.index(b"\x00", pos)
+            pos = z + 1
+        # skip unknown 4 bytes
+        pos += 4
+        if pos + 4 > len(data):
+            return data
+        size = int.from_bytes(data[pos:pos+4], "little")
+        pos += 4
+        if pos + size <= len(data):
+            return data[pos:pos+size]
+    except Exception:
+        pass
+    return data
+
+def extract_embed_file(target: Union[bytes, bytearray]) -> List[Tuple[str, bytes]]:
+    """
+    Only extract the 'first layer' of embedding, returning raw (filename, bytes).
+    """
+    top = bytes(target)
+    head = top[:8]
+    out: List[Tuple[str, bytes]] = []
+    seen = set()
+
+    def push(b: bytes, name_hint: str = ""):
+        h10 = _sha10(b)
+        if h10 in seen:
+            return
+        seen.add(h10)
+        ext = _guess_ext(b)
+        # If name_hint has an extension use its basename; else fallback to guessed ext
+        if "." in name_hint:
+            fname = name_hint.split("/")[-1]
+        else:
+            fname = f"{h10}{ext}"
+        out.append((fname, b))
+
+    # OOXML/ZIP container (docx/xlsx/pptx)
+    if _is_zip(head):
+        try:
+            with zipfile.ZipFile(io.BytesIO(top), "r") as z:
+                embed_dirs = (
+                    "word/embeddings/", "word/objects/", "word/activex/",
+                    "xl/embeddings/", "ppt/embeddings/"
+                )
+                for name in z.namelist():
+                    low = name.lower()
+                    if any(low.startswith(d) for d in embed_dirs):
+                        try:
+                            b = z.read(name)
+                            push(b, name)
+                        except Exception:
+                            pass
+        except Exception:
+            pass
+        return out
+
+    # OLE container (doc/ppt/xls)
+    if _is_ole(head):
+        try:
+            with olefile.OleFileIO(io.BytesIO(top)) as ole:
+                for entry in ole.listdir():
+                    p = "/".join(entry)
+                    try:
+                        data = ole.openstream(entry).read()
+                    except Exception:
+                        continue
+                    if not data:
+                        continue
+                    if "Ole10Native" in p or "ole10native" in p.lower():
+                        data = _extract_ole10native_payload(data)
+                    push(data, p)
+        except Exception:
+            pass
+        return out
+
+    return out

api/utils/health.py

@@ -0,0 +1,104 @@
+from timeit import default_timer as timer
+
+from api import settings
+from api.db.db_models import DB
+from rag.utils.redis_conn import REDIS_CONN
+from rag.utils.storage_factory import STORAGE_IMPL
+
+
+def _ok_nok(ok: bool) -> str:
+    return "ok" if ok else "nok"
+
+
+def check_db() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        # lightweight probe; works for MySQL/Postgres
+        DB.execute_sql("SELECT 1")
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_redis() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        ok = bool(REDIS_CONN.health())
+        return ok, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_doc_engine() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        meta = settings.docStoreConn.health()
+        # treat any successful call as ok
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", **(meta or {})}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_storage() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        STORAGE_IMPL.health()
+        return True, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def check_chat() -> tuple[bool, dict]:
+    st = timer()
+    try:
+        cfg = getattr(settings, "CHAT_CFG", None)
+        ok = bool(cfg and cfg.get("factory"))
+        return ok, {"elapsed": f"{(timer() - st) * 1000.0:.1f}"}
+    except Exception as e:
+        return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
+
+
+def run_health_checks() -> tuple[dict, bool]:
+    result: dict[str, str | dict] = {}
+
+    db_ok, db_meta = check_db()
+    chat_ok, chat_meta = check_chat()
+
+    result["db"] = _ok_nok(db_ok)
+    if not db_ok:
+        result.setdefault("_meta", {})["db"] = db_meta
+
+    result["chat"] = _ok_nok(chat_ok)
+    if not chat_ok:
+        result.setdefault("_meta", {})["chat"] = chat_meta
+
+    # Optional probes (do not change minimal contract but exposed for observability)
+    try:
+        redis_ok, redis_meta = check_redis()
+        result["redis"] = _ok_nok(redis_ok)
+        if not redis_ok:
+            result.setdefault("_meta", {})["redis"] = redis_meta
+    except Exception:
+        result["redis"] = "nok"
+
+    try:
+        doc_ok, doc_meta = check_doc_engine()
+        result["doc_engine"] = _ok_nok(doc_ok)
+        if not doc_ok:
+            result.setdefault("_meta", {})["doc_engine"] = doc_meta
+    except Exception:
+        result["doc_engine"] = "nok"
+
+    try:
+        sto_ok, sto_meta = check_storage()
+        result["storage"] = _ok_nok(sto_ok)
+        if not sto_ok:
+            result.setdefault("_meta", {})["storage"] = sto_meta
+    except Exception:
+        result["storage"] = "nok"
+
+    all_ok = (result.get("db") == "ok") and (result.get("chat") == "ok")
+    result["status"] = "ok" if all_ok else "nok"
+    return result, all_ok
+
+

api/utils/health_utils.py

@@ -13,14 +13,17 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
-
+import os
-
+import requests
 from timeit import default_timer as timer
 
 from api import settings
 from api.db.db_models import DB
+from rag import settings as rag_settings
 from rag.utils.redis_conn import REDIS_CONN
 from rag.utils.storage_factory import STORAGE_IMPL
+from rag.utils.es_conn import ESConnection
+from rag.utils.infinity_conn import InfinityConnection
 
 
 def _ok_nok(ok: bool) -> str:
@@ -65,6 +68,96 @@ def check_storage() -> tuple[bool, dict]:
         return False, {"elapsed": f"{(timer() - st) * 1000.0:.1f}", "error": str(e)}
 
 
+def get_es_cluster_stats() -> dict:
+    doc_engine = os.getenv('DOC_ENGINE', 'elasticsearch')
+    if doc_engine != 'elasticsearch':
+        raise Exception("Elasticsearch is not in use.")
+    try:
+        return {
+            "status": "alive",
+            "message": ESConnection().get_cluster_stats()
+        }
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
+
+
+def get_infinity_status():
+    doc_engine = os.getenv('DOC_ENGINE', 'elasticsearch')
+    if doc_engine != 'infinity':
+        raise Exception("Infinity is not in use.")
+    try:
+        return {
+            "status": "alive",
+            "message": InfinityConnection().health()
+        }
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
+
+
+def get_mysql_status():
+    try:
+        cursor = DB.execute_sql("SHOW PROCESSLIST;")
+        res_rows = cursor.fetchall()
+        headers = ['id', 'user', 'host', 'db', 'command', 'time', 'state', 'info']
+        cursor.close()
+        return {
+            "status": "alive",
+            "message": [dict(zip(headers, r)) for r in res_rows]
+        }
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
+
+
+def check_minio_alive():
+    start_time = timer()
+    try:
+        response = requests.get(f'http://{rag_settings.MINIO["host"]}/minio/health/live')
+        if response.status_code == 200:
+            return {"status": "alive", "message": f"Confirm elapsed: {(timer() - start_time) * 1000.0:.1f} ms."}
+        else:
+            return {"status": "timeout", "message": f"Confirm elapsed: {(timer() - start_time) * 1000.0:.1f} ms."}
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
+
+
+def get_redis_info():
+    try:
+        return {
+            "status": "alive",
+            "message": REDIS_CONN.info()
+        }
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
+
+
+def check_ragflow_server_alive():
+    start_time = timer()
+    try:
+        response = requests.get(f'http://{settings.HOST_IP}:{settings.HOST_PORT}/v1/system/ping')
+        if response.status_code == 200:
+            return {"status": "alive", "message": f"Confirm elapsed: {(timer() - start_time) * 1000.0:.1f} ms."}
+        else:
+            return {"status": "timeout", "message": f"Confirm elapsed: {(timer() - start_time) * 1000.0:.1f} ms."}
+    except Exception as e:
+        return {
+            "status": "timeout",
+            "message": f"error: {str(e)}",
+        }
 
 
 def run_health_checks() -> tuple[dict, bool]:
@@ -99,9 +192,7 @@ def run_health_checks() -> tuple[dict, bool]:
     except Exception:
         result["storage"] = "nok"
 
-
+    all_ok = (result.get("db") == "ok") and (result.get("redis") == "ok") and (result.get("doc_engine") == "ok") and (
-    all_ok = (result.get("db") == "ok") and (result.get("redis") == "ok") and (result.get("doc_engine") == "ok") and (result.get("storage") == "ok")
+                result.get("storage") == "ok")
     result["status"] = "ok" if all_ok else "nok"
     return result, all_ok
-
-

api/utils/web_utils.py

@@ -24,6 +24,7 @@ from urllib.parse import urlparse
 from api.apps import smtp_mail_server
 from flask_mail import Message
 from flask import render_template_string
+from api.utils.email_templates import EMAIL_TEMPLATES
 from selenium import webdriver
 from selenium.common.exceptions import TimeoutException
 from selenium.webdriver.chrome.options import Options
@@ -34,6 +35,12 @@ from selenium.webdriver.support.ui import WebDriverWait
 from webdriver_manager.chrome import ChromeDriverManager
 
 
+OTP_LENGTH = 8
+OTP_TTL_SECONDS = 5 * 60
+ATTEMPT_LIMIT = 5
+ATTEMPT_LOCK_SECONDS = 30 * 60
+RESEND_COOLDOWN_SECONDS = 60
+
 
 CONTENT_TYPE_MAP = {
     # Office
@@ -178,24 +185,49 @@ def get_float(req: dict, key: str, default: float | int = 10.0) -> float:
         return default
 
 
-INVITE_EMAIL_TMPL = """
+def send_email_html(subject: str, to_email: str, template_key: str, **context):
-<p>Hi {{email}},</p>
+    """Generic HTML email sender using shared templates.
-<p>{{inviter}} has invited you to join their team (ID: {{tenant_id}}).</p>
+    template_key must exist in EMAIL_TEMPLATES.
-<p>Click the link below to complete your registration:<br>
+    """
-<a href="{{invite_url}}">{{invite_url}}</a></p>
+    from api.apps import app
-<p>If you did not request this, please ignore this email.</p>
+    tmpl = EMAIL_TEMPLATES.get(template_key)
-"""
+    if not tmpl:
+        raise ValueError(f"Unknown email template: {template_key}")
+    with app.app_context():
+        msg = Message(subject=subject, recipients=[to_email])
+        msg.html = render_template_string(tmpl, **context)
+        smtp_mail_server.send(msg)
+
 
 def send_invite_email(to_email, invite_url, tenant_id, inviter):
-    from api.apps import  app
+    # Reuse the generic HTML sender with 'invite' template
-    with app.app_context():
+    send_email_html(
-        msg = Message(subject="RAGFlow Invitation",
+        subject="RAGFlow Invitation",
-                      recipients=[to_email])
+        to_email=to_email,
-        msg.html = render_template_string(
+        template_key="invite",
-            INVITE_EMAIL_TMPL,
+        email=to_email,
-            email=to_email,
+        invite_url=invite_url,
-            invite_url=invite_url,
+        tenant_id=tenant_id,
-            tenant_id=tenant_id,
+        inviter=inviter,
-            inviter=inviter,
+    )
-        )
+
-        smtp_mail_server.send(msg)
+
+def otp_keys(email: str):
+    email = (email or "").strip().lower()
+    return (
+        f"otp:{email}",
+        f"otp_attempts:{email}",
+        f"otp_last_sent:{email}",
+        f"otp_lock:{email}",
+    )
+
+
+def hash_code(code: str, salt: bytes) -> str:
+    import hashlib
+    import hmac 
+    return hmac.new(salt, (code or "").encode("utf-8"), hashlib.sha256).hexdigest()
+    
+
+def captcha_key(email: str) -> str:
+    return f"captcha:{email}"
+    

conf/infinity_mapping.json

@@ -31,7 +31,6 @@
 	"entities_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
 	"pagerank_fea": {"type": "integer", "default":  0},
 	"tag_feas": {"type": "varchar", "default": "", "analyzer": "rankfeatures"},
-
 	"from_entity_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
 	"to_entity_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
 	"entity_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
@@ -39,6 +38,6 @@
 	"source_id": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
 	"n_hop_with_weight": {"type": "varchar", "default": ""},
 	"removed_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
-
+	"doc_type_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"},
-	"doc_type_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"}
+	"toc_kwd": {"type": "varchar", "default": "", "analyzer": "whitespace-#"}
 }

conf/llm_factories.json

@@ -803,6 +803,12 @@
                     "tags": "TEXT EMBEDDING",
                     "max_tokens": 512,
                     "model_type": "embedding"
+                },
+                {
+                    "llm_name": "glm-asr",
+                    "tags": "SPEECH2TEXT",
+                    "max_tokens": 4096,
+                    "model_type": "speech2text"
                 }
             ]
         },
@@ -965,31 +971,9 @@
         {
             "name": "VolcEngine",
             "logo": "",
-            "tags": "LLM, TEXT EMBEDDING",
+            "tags": "LLM, TEXT EMBEDDING, IMAGE2TEXT",
             "status": "1",
-            "llm": [
+            "llm": []
-                {
-                    "llm_name": "Doubao-pro-128k",
-                    "tags": "LLM,CHAT,128k",
-                    "max_tokens": 131072,
-                    "model_type": "chat",
-                    "is_tools": true
-                },
-                {
-                    "llm_name": "Doubao-pro-32k",
-                    "tags": "LLM,CHAT,32k",
-                    "max_tokens": 32768,
-                    "model_type": "chat",
-                    "is_tools": true
-                },
-                {
-                    "llm_name": "Doubao-pro-4k",
-                    "tags": "LLM,CHAT,4k",
-                    "max_tokens": 4096,
-                    "model_type": "chat",
-                    "is_tools": true
-                }
-            ]
         },
         {
             "name": "BaiChuan",
@@ -1361,35 +1345,35 @@
                     "llm_name": "gemini-2.5-flash",
                     "tags": "LLM,CHAT,1024K,IMAGE2TEXT",
                     "max_tokens": 1048576,
-                    "model_type": "chat",
+                    "model_type": "image2text",
                     "is_tools": true
                 },
                 {
                     "llm_name": "gemini-2.5-pro",
                     "tags": "LLM,CHAT,IMAGE2TEXT,1024K",
                     "max_tokens": 1048576,
-                    "model_type": "chat",
+                    "model_type": "image2text",
                     "is_tools": true
                 },
                 {
                     "llm_name": "gemini-2.5-flash-lite",
                     "tags": "LLM,CHAT,1024K,IMAGE2TEXT",
                     "max_tokens": 1048576,
-                    "model_type": "chat",
+                    "model_type": "image2text",
                     "is_tools": true
                 },
                 {
                     "llm_name": "gemini-2.0-flash",
                     "tags": "LLM,CHAT,1024K",
                     "max_tokens": 1048576,
-                    "model_type": "chat",
+                    "model_type": "image2text",
                     "is_tools": true
                 },
                 {
                     "llm_name": "gemini-2.0-flash-lite",
                     "tags": "LLM,CHAT,1024K",
                     "max_tokens": 1048576,
-                    "model_type": "chat",
+                    "model_type": "image2text",
                     "is_tools": true
                 },
                 {
@@ -2816,6 +2800,13 @@
             "tags": "LLM,TEXT EMBEDDING,TEXT RE-RANK,IMAGE2TEXT",
             "status": "1",
             "llm": [
+                {
+                    "llm_name":"THUDM/GLM-4.1V-9B-Thinking",
+                    "tags":"LLM,CHAT,IMAGE2TEXT, 64k",
+                    "max_tokens":64000,
+                    "model_type":"chat",
+                    "is_tools": false
+                },
                 {
                     "llm_name": "Qwen/Qwen3-Embedding-8B",
                     "tags": "TEXT EMBEDDING,TEXT RE-RANK,32k",
@@ -2996,7 +2987,7 @@
                     "tags": "LLM,CHAT,IMAGE2TEXT,32k",
                     "max_tokens": 32000,
                     "model_type": "image2text",
-                    "is_tools": true
+                    "is_tools": false
                 },
                 {
                     "llm_name": "THUDM/GLM-Z1-32B-0414",
@@ -3145,13 +3136,6 @@
                     "model_type": "chat",
                     "is_tools": true
                 },
-                {
-                    "llm_name": "Qwen/Qwen2-1.5B-Instruct",
-                    "tags": "LLM,CHAT,32k",
-                    "max_tokens": 32000,
-                    "model_type": "chat",
-                    "is_tools": true
-                },
                 {
                     "llm_name": "Pro/Qwen/Qwen2.5-Coder-7B-Instruct",
                     "tags": "LLM,CHAT,32k",
@@ -3159,13 +3143,6 @@
                     "model_type": "chat",
                     "is_tools": false
                 },
-                {
-                    "llm_name": "Pro/Qwen/Qwen2-VL-7B-Instruct",
-                    "tags": "LLM,CHAT,IMAGE2TEXT,32k",
-                    "max_tokens": 32000,
-                    "model_type": "image2text",
-                    "is_tools": false
-                },
                 {
                     "llm_name": "Pro/Qwen/Qwen2.5-7B-Instruct",
                     "tags": "LLM,CHAT,32k",
@@ -3533,6 +3510,13 @@
                     "model_type": "chat",
                     "is_tools": true
                 },
+                {
+                    "llm_name": "claude-sonnet-4-5-20250929",
+                    "tags": "LLM,CHAT,IMAGE2TEXT,200k",
+                    "max_tokens": 204800,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
                 {
                     "llm_name": "claude-sonnet-4-20250514",
                     "tags": "LLM,CHAT,IMAGE2TEXT,200k",
@@ -4862,8 +4846,282 @@
                     "max_tokens": 8000,
                     "model_type": "chat",
                     "is_tools": true
+                },
+                {
+                    "llm_name": "LongCat-Flash-Thinking",
+                    "tags": "LLM,CHAT,8000",
+                    "max_tokens": 8000,
+                    "model_type": "chat",
+                    "is_tools": true
+                }
+            ]
+        },
+        {
+            "name": "DeerAPI",
+            "logo": "",
+            "tags": "LLM,TEXT EMBEDDING,IMAGE2TEXT",
+            "status": "1",
+            "llm": [
+                {
+                    "llm_name": "gpt-5-chat-latest",
+                    "tags": "LLM,CHAT,400k",
+                    "max_tokens": 400000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "chatgpt-4o-latest",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-5-mini",
+                    "tags": "LLM,CHAT,400k",
+                    "max_tokens": 400000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-5-nano",
+                    "tags": "LLM,CHAT,400k",
+                    "max_tokens": 400000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-5",
+                    "tags": "LLM,CHAT,400k",
+                    "max_tokens": 400000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-4.1-mini",
+                    "tags": "LLM,CHAT,1M",
+                    "max_tokens": 1047576,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-4.1-nano",
+                    "tags": "LLM,CHAT,1M",
+                    "max_tokens": 1047576,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-4.1",
+                    "tags": "LLM,CHAT,1M",
+                    "max_tokens": 1047576,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gpt-4o-mini",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "o4-mini-2025-04-16",
+                    "tags": "LLM,CHAT,200k",
+                    "max_tokens": 200000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "o3-pro-2025-06-10",
+                    "tags": "LLM,CHAT,200k",
+                    "max_tokens": 200000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-opus-4-1-20250805",
+                    "tags": "LLM,CHAT,200k,IMAGE2TEXT",
+                    "max_tokens": 200000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-opus-4-1-20250805-thinking",
+                    "tags": "LLM,CHAT,200k,IMAGE2TEXT",
+                    "max_tokens": 200000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-sonnet-4-20250514",
+                    "tags": "LLM,CHAT,200k,IMAGE2TEXT",
+                    "max_tokens": 200000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-sonnet-4-20250514-thinking",
+                    "tags": "LLM,CHAT,200k,IMAGE2TEXT",
+                    "max_tokens": 200000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-3-7-sonnet-latest",
+                    "tags": "LLM,CHAT,200k",
+                    "max_tokens": 200000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "claude-3-5-haiku-latest",
+                    "tags": "LLM,CHAT,200k",
+                    "max_tokens": 200000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gemini-2.5-pro",
+                    "tags": "LLM,CHAT,1M,IMAGE2TEXT",
+                    "max_tokens": 1000000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gemini-2.5-flash",
+                    "tags": "LLM,CHAT,1M,IMAGE2TEXT",
+                    "max_tokens": 1000000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gemini-2.5-flash-lite",
+                    "tags": "LLM,CHAT,1M,IMAGE2TEXT",
+                    "max_tokens": 1000000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "gemini-2.0-flash",
+                    "tags": "LLM,CHAT,1M,IMAGE2TEXT",
+                    "max_tokens": 1000000,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "grok-4-0709",
+                    "tags": "LLM,CHAT,131k",
+                    "max_tokens": 131072,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "grok-3",
+                    "tags": "LLM,CHAT,131k",
+                    "max_tokens": 131072,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "grok-3-mini",
+                    "tags": "LLM,CHAT,131k",
+                    "max_tokens": 131072,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "grok-2-image-1212",
+                    "tags": "LLM,CHAT,32k,IMAGE2TEXT",
+                    "max_tokens": 32768,
+                    "model_type": "image2text",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "deepseek-v3.1",
+                    "tags": "LLM,CHAT,64k",
+                    "max_tokens": 64000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "deepseek-v3",
+                    "tags": "LLM,CHAT,64k",
+                    "max_tokens": 64000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "deepseek-r1-0528",
+                    "tags": "LLM,CHAT,164k",
+                    "max_tokens": 164000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "deepseek-chat",
+                    "tags": "LLM,CHAT,32k",
+                    "max_tokens": 32000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "deepseek-reasoner",
+                    "tags": "LLM,CHAT,64k",
+                    "max_tokens": 64000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "qwen3-30b-a3b",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "qwen3-coder-plus-2025-07-22",
+                    "tags": "LLM,CHAT,128k",
+                    "max_tokens": 128000,
+                    "model_type": "chat",
+                    "is_tools": true
+                },
+                {
+                    "llm_name": "text-embedding-ada-002",
+                    "tags": "TEXT EMBEDDING,8K",
+                    "max_tokens": 8191,
+                    "model_type": "embedding",
+                    "is_tools": false
+                },
+                {
+                    "llm_name": "text-embedding-3-small",
+                    "tags": "TEXT EMBEDDING,8K",
+                    "max_tokens": 8191,
+                    "model_type": "embedding",
+                    "is_tools": false
+                },
+                {
+                    "llm_name": "text-embedding-3-large",
+                    "tags": "TEXT EMBEDDING,8K",
+                    "max_tokens": 8191,
+                    "model_type": "embedding",
+                    "is_tools": false
+                },
+                {
+                    "llm_name": "whisper-1",
+                    "tags": "SPEECH2TEXT",
+                    "max_tokens": 26214400,
+                    "model_type": "speech2text",
+                    "is_tools": false
+                },
+                {
+                    "llm_name": "tts-1",
+                    "tags": "TTS",
+                    "max_tokens": 2048,
+                    "model_type": "tts",
+                    "is_tools": false
                 }
             ]
         }
     ]
-}
+}