mirror of
https://github.com/langflow-ai/langflow.git
synced 2026-07-24 00:39:35 +08:00
43253011ec
* checkout wxo deployment adapter implementation * checkout wxo deps and flow reqs impl * clean up / minor refactor with updated tests * major refactor: split up the implementation into folders and files * clean up logic * refactor and clean up of execution logic, improve type safety of "retry_create" helper, and remove dead code * remove unused wxo service file * remove "provider_name" arg and references * fix: harden watsonx orchestrate deployment adapter for PR readiness - Fix rollback AttributeError when agent_create_response is None - Fix NoneType access on params in list() when called without params - Fix inconsistent error types: use DeploymentNotFoundError in get/update - Fix typo "occured" -> "occurred" in all error prefix messages - Fix variable shadowing of fastapi.status in get_status() - Fix pre-existing test bugs (wrong exception types, stale method refs) - Fix e2e monkey-patching of non-existent service methods - Add structured logging to create, delete, retry, and rollback flows - Add jitter to retry backoff to avoid thundering herd - Add __repr__ to WxOCredentials that fully masks the API key - Extract hard-coded LLM model string to DEFAULT_WXO_AGENT_LLM constant - Remove commented-out snapshot update code - Expand test coverage from 24 to 75 tests covering retry logic, service methods, client auth, utilities, execution/status helpers, and artifact roundtrip * fix(deployment): fix bugs and harden watsonx orchestrate adapter - Fix update method silently discarding changes instead of sending them to the WXO API - Wrap all synchronous SDK calls in asyncio.to_thread to avoid blocking the event loop - Add error handling to get_status (was propagating raw exceptions) - Use DeploymentType enum values instead of raw strings for SUPPORTED_ADAPTER_DEPLOYMENT_TYPES - Fix type annotations in list method (list -> set) - Fix typo in comment (reccomend -> recommend) - Remove dead code: extract_agent_connection_ids, require_exclusive_resource, require_non_empty_string, resolve_health_environment_id, fetch_agent_release_status, normalize_release_status, resolve_lfx_runner_requirement, _pin_requirement_name, sync_langflow_tool_connections, create_langflow_flow_tool, resolve_snapshot_connections, and unused constants - Make assert_create_resources_available and validate_connection async - Make create_agent_run and get_agent_run async - Add tests for list_types, get_status error handling, and update side-effects - Update existing tests for async function signatures * actually remove the dead code from tools.py * properly await "validate_connection" * update e2e test file * add more scenarios to e2e test runner * refactor(watsonx-orchestrate): improve WxOClient encapsulation and error messages Encapsulate SDK private method access and endpoint paths inside WxOClient wrappers so callers never handle raw paths or touch internal _get/_post methods. Route all private HTTP calls through a single _base client auto-created in __post_init__, removing the externally constructed `base` field. Additionally fix double-period and redundant text in error messages produced by the ErrorPrefix + handler pattern, and update E2E/unit assertions to match the sanitised error output. Changes: - types.py: bake endpoint paths into wrapper signatures (post_run, get_run, upload_tool_artifact, get_agents_raw); auto-create _base via __post_init__; remove base from constructor - client.py: drop BaseWXOClient import and base= constructor arg - core/execution.py: pass run_id / query_suffix instead of raw paths - core/tools.py: pass tool_id instead of raw path - service.py: fix double-period in ErrorPrefix interpolation; remove redundant restatements in generic except handlers - tests: update _with_wxo_wrappers helper and test doubles to use _base; update FakeBaseClient._get to accept params - e2e: update rollback scenario detail_contains to match sanitised msg * skip import and tests of wxo adapter if current env is running python 3.10 * clarify random prefix / retry behavior * implement comments * fix: align watsonx adapter with updated deployment schema - Add `deployment_type` parameter to `get`, `update`, `redeploy`, `duplicate`, `delete`, `get_status`, and `undeploy_deployment` in WatsonxOrchestrateDeploymentService to match the updated BaseDeploymentService ABC - Update e2e script to use renamed `add_ids` field on SnapshotDeploymentBindingUpdate * remove non-interface method undeploy_deployment * implement listing configs and snapshots * add update implementation and improve http->deployment error translation and add tests * improve exception handling * checkout payload slot work * custom payload schema for update * new update implementation * stop passing client cache to helpers * add docs for ordereduniquests * improve import patterns and document future risks for wxo dependencies * remove global-variable prefixing of flows * ref: harden typing, DRY helpers, and correctness fixes - Replace `db: Any` with `db: AsyncSession` across client, config, and update_helpers modules - Extract `_ensure_dict` / `ensure_langflow_connections_binding` helpers to eliminate repeated nested-dict safety logic in tools.py and update_helpers.py, with documentation explaining why malformed API payloads are silently corrected rather than rejected - Use `PayloadSlot.parse()` for update payload validation instead of standalone helper; remove `parse_provider_update_payload` - Fix lambda late-binding with default-argument captures in service.py - Use `zip(strict=True)` in update_helpers for defensive mismatch detection - Replace O(n²) duplicate detection with `collections.Counter` in payloads.py - Simplify `dedupe_list` to `list(dict.fromkeys(...))` - Move type-annotation-only imports into TYPE_CHECKING blocks in types.py and config.py - Introduce `UPDATE_MAX_RETRIES` as a distinct constant from `CREATE_MAX_RETRIES` - Fix "watsonX" → "watsonx" casing in error messages - Clarify `get_status` docstring and `validate_connection` error message - Add TODO(deployments-cache) for client cache invalidation on credential updates * ref: drop client cache and adopt typed deployment context across the wxo adapter path, add request-context memoization with strict mixed-context guards, and lazily initialize wxo SDK clients. Add safer credential handling by storing only authenticators in WxOCredentials. Also improve provider error-detail extraction/messages and make the direct wxo E2E conflict scenario diagnostics and expectations more resilient. * use explicit naming for context class; providerId * fix error handling, retry safety, and exception diagnostics in wxo adapter - Raise DeploymentError on empty API responses instead of fabricating fake success in create_agent_run_result and get_agent_run - Elevate rollback failure logging from WARNING to ERROR for alerting - Apply retryable filter to retry_rollback to skip 401/403/409/422 - Preserve exception chains (from exc) instead of suppressing (from None) across service.py, utils.py, execution.py, and update_helpers.py - Broaden credential resolution catch to handle arbitrary DB exceptions - Separate status code dispatch from string heuristics in raise_for_status_and_detail to prevent misclassification - Add tests for all behavioral changes * [autofix.ci] apply automated fixes * fix: harden wxO adapter safety, immutability, and error diagnostics - Make WxOClient and WxOCredentials frozen dataclasses with eager SDK client initialization to eliminate thread-safety races from asyncio.to_thread workers and prevent post-construction mutation - Move instance_url validation and normalization into type __post_init__ - Raise DeploymentError on missing run_id instead of returning partial success with execution_id=None - Preserve exception chains (from exc) instead of suppressing (from None) across create, update, and delete service methods - Add warning logs when _ensure_dict replaces non-dict binding values and when _resolve_lfx_requirement falls back to minimum version - Initialize derived_spec before try block to prevent potential NameError - Log all ToolUploadBatchError errors before re-raising the first - Change SUPPORTED_ADAPTER_DEPLOYMENT_TYPES from mutable set to frozenset - Add tests for 409/422 error mapping in create, unsupported deployment type rejection, empty update rejection, zip artifact extraction paths, validate_connection negative paths, missing run_id, multiple deployment ID rejection, exception chain preservation, and _ensure_dict warning * [autofix.ci] apply automated fixes * fix ruff errors and and todo for status method * fix mypy errors * [autofix.ci] apply automated fixes --------- Co-authored-by: Jordan Frazier <jordan.frazier@datastax.com> Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Langflow is a powerful platform for building and deploying AI-powered agents and workflows. It provides developers with both a visual authoring experience and built-in API and MCP servers that turn every workflow into a tool that can be integrated into applications built on any framework or stack. Langflow comes with batteries included and supports all major LLMs, vector databases and a growing library of AI tools.
✨ Highlight features
- Visual builder interface to quickly get started and iterate.
- Source code access lets you customize any component using Python.
- Interactive playground to immediately test and refine your flows with step-by-step control.
- Multi-agent orchestration with conversation management and retrieval.
- Deploy as an API or export as JSON for Python apps.
- Deploy as an MCP server and turn your flows into tools for MCP clients.
- Observability with LangSmith, LangFuse and other integrations.
- Enterprise-ready security and scalability.
🖥️ Langflow Desktop
Langflow Desktop is the easiest way to get started with Langflow. All dependencies are included, so you don't need to manage Python environments or install packages manually. Available for Windows and macOS.
⚡️ Quickstart
Install locally (recommended)
Requires Python 3.10–3.13 and uv (recommended package manager).
Install
From a fresh directory, run:
uv pip install langflow -U
The latest Langflow package is installed. For more information, see Install and run the Langflow OSS Python package.
Run
To start Langflow, run:
uv run langflow run
Langflow starts at http://127.0.0.1:7860.
That's it! You're ready to build with Langflow! 🎉
📦 Other install options
Run from source
If you've cloned this repository and want to contribute, run this command from the repository root:
make run_cli
For more information, see DEVELOPMENT.md.
Docker
Start a Langflow container with default settings:
docker run -p 7860:7860 langflowai/langflow:latest
Langflow is available at http://localhost:7860/. For configuration options, see the Docker deployment guide.
Caution
- Users must update to Langflow >= 1.7.1 to protect against CVE-2025-68477 and CVE-2025-68478.
- Langflow version 1.7.0 has a critical bug where persisted state (flows, projects, and global variables) cannot be found when upgrading. Version 1.7.0 was yanked and replaced with version 1.7.1, which includes a fix for this bug. DO NOT upgrade to version 1.7.0. Instead, upgrade directly to version 1.7.1.
- Langflow versions 1.6.0 through 1.6.3 have a critical bug where
.envfiles are not read, potentially causing security vulnerabilities. DO NOT upgrade to these versions if you use.envfiles for configuration. Instead, upgrade to 1.6.4, which includes a fix for this bug.- Windows users of Langflow Desktop should not use the in-app update feature to upgrade to Langflow version 1.6.0. For upgrade instructions, see Windows Desktop update issue.
- Users must update to Langflow >= 1.3 to protect against CVE-2025-3248
- Users must update to Langflow >= 1.5.1 to protect against CVE-2025-57760
For security information, see our Security Policy and Security Advisories.
🚀 Deployment
Langflow is completely open source and you can deploy it to all major deployment clouds. To learn how to deploy Langflow, see our Langflow deployment guides.
⭐ Stay up-to-date
Star Langflow on GitHub to be instantly notified of new releases.
👋 Contribute
We welcome contributions from developers of all levels. If you'd like to contribute, please check our contributing guidelines and help make Langflow more accessible.