Commit Graph

  • 416c2665a7 fix: Rebuild AgentResponse model in multi-turn A2A flows to support skill IDs Devin AI 2025-11-12 18:02:58 +00:00
  • d141078e72 fix: Allow A2A agents to be identified by skill ID in addition to endpoint URL Devin AI 2025-11-12 17:51:48 +00:00
  • c205d2e8de feat: implement before and after LLM call hooks in CrewAgentExecutor (#3893) Lorenze Jay 2025-11-12 08:38:13 -08:00
  • 8b83bf3e54 chore: remove duplication in azure client Greyson LaLonde 2025-11-11 18:01:27 -05:00
  • 93f1fbd75e chore: move api key validation to base Greyson LaLonde 2025-11-11 17:46:26 -05:00
  • 0803318002 chore: improve typing Greyson LaLonde 2025-11-11 17:37:08 -05:00
  • fcb5b19b2e Enhance schema description of QdrantVectorSearchTool (#3891) Daniel Barreto 2025-11-11 19:33:33 -03:00
  • 6fb13ee3e0 chore: fix attr ref Greyson LaLonde 2025-11-11 00:02:34 -05:00
  • 1b7f44dc34 Fix GPT-5 tool calling format incompatibility devin/1762863320-fix-gpt5-tool-calling-format Devin AI 2025-11-11 12:23:39 +00:00
  • 389d68efb4 Fix hierarchical delegation bug where manager couldn't delegate to all crew agents devin/1762858722-fix-hierarchical-delegation-bug-3887 Devin AI 2025-11-11 11:07:22 +00:00
  • e9b3fded3a Fix AIMindTool import error with minds_sdk 2.0.0 devin/1762848818-fix-aimind-tool-import Devin AI 2025-11-11 08:25:07 +00:00
  • e5a117c4ca Re-run CI for cancelled tests devin/1762847531-fix-tool-import-from-crewai-tools Devin AI 2025-11-11 08:02:47 +00:00
  • 230af2749b Fix #3883: Add compatibility alias for tool decorator in crewai_tools Devin AI 2025-11-11 07:58:29 +00:00
  • 67e39073c7 Merge branch 'gl/chore/use-base-model-for-llms' of https://github.com/crewAIInc/crewAI into gl/chore/use-base-model-for-llms Greyson LaLonde 2025-11-10 23:51:56 -05:00
  • 722d316824 chore: continue refactoring llms to base models Greyson LaLonde 2025-11-10 16:05:23 -05:00
  • a824d52e5e Merge branch 'main' into gl/chore/use-base-model-for-llms Greyson LaLonde 2025-11-10 23:40:41 -05:00
  • d8fe83f76c chore: continue refactoring llms to base models Greyson LaLonde 2025-11-10 16:05:23 -05:00
  • 01f0111d52 dependabot.yml creation (#3868) Rip&Tear 2025-11-11 12:14:16 +08:00
  • 6b52587c67 feat: expose messages to TaskOutput and LiteAgentOutputs (#3880) Lorenze Jay 2025-11-10 17:38:30 -08:00
  • 629f7f34ce docs: enhance task guardrail documentation with LLM-based validation support (#3879) Lorenze Jay 2025-11-10 15:35:42 -08:00
  • 46785adf58 chore: refactor llms to base models Greyson LaLonde 2025-11-10 14:22:09 -05:00
  • d25bea781d fix: Resolve test failures in HTTP MCP support tests devin/1762786788-add-http-support-for-mcp-servers Devin AI 2025-11-10 15:22:50 +00:00
  • 6063677c01 fix: Correct MCPToolWrapper patch path in fragment filtering test Devin AI 2025-11-10 15:17:30 +00:00
  • cfaa44012f fix: Update warning log tests to use custom Logger monkeypatch Devin AI 2025-11-10 15:10:59 +00:00
  • 697182b0ef feat: Add HTTP support for local MCP servers (#3876) Devin AI 2025-11-10 15:05:30 +00:00
  • eee3556b52 Revert to pytest-recording to fix CI --block-network flag devin/1762773888-fix-agent-thought-action-leak Devin AI 2025-11-10 11:42:40 +00:00
  • efb005afaf Fix agent output sanitization to prevent internal ReAct fields from leaking Devin AI 2025-11-10 11:32:48 +00:00
  • ce8462f2f3 Fix lint error and test failures devin/1762750602-fix-kickoff-for-each-hang Devin AI 2025-11-10 05:08:36 +00:00
  • f1bc535ff8 Fix issue #3871: kickoff_for_each() hangs after completion Devin AI 2025-11-10 05:01:20 +00:00
  • c2d4073415 Potential fix for code scanning alert no. 23: Incomplete URL substring sanitization Rip&Tear 2025-11-08 21:34:00 +08:00
  • f557334c01 Potential fix for code scanning alert no. 26: Incomplete URL substring sanitization Rip&Tear 2025-11-08 21:31:40 +08:00
  • 3ab991d206 Potential fix for code scanning alert no. 25: Incomplete URL substring sanitization Rip&Tear 2025-11-08 21:18:30 +08:00
  • e3a0cda16c Potential fix for code scanning alert no. 27: Incomplete URL substring sanitization Rip&Tear 2025-11-08 21:16:07 +08:00
  • 9e75055c3e chore: adjust cassette dir and regenerate openai gl/chore/pytest-vcr-cassette-updates Greyson LaLonde 2025-11-08 02:26:18 -05:00
  • b9bc5714c3 chore: delete all cassettes and update conftest Greyson LaLonde 2025-11-08 01:54:25 -05:00
  • 996c53228b chore: organize pyproject, update pytest ini Greyson LaLonde 2025-11-08 01:50:29 -05:00
  • d0d22038f5 chore: remove telemetry mocks Greyson LaLonde 2025-11-07 17:28:32 -05:00
  • 2ba0434eb8 Merge branch 'main' into gl/chore/pytest-vcr-cassette-updates Greyson LaLonde 2025-11-07 17:09:45 -05:00
  • cad3251721 chore: delete cassettes and make vcr_config session scope Greyson Lalonde 2025-11-07 17:07:11 -05:00
  • 19f4ef0982 chore: remove decorator kwarg to ensure global conftest applies Greyson Lalonde 2025-11-07 17:01:30 -05:00
  • 54f8cc685f chore: update pytest-vcr config Greyson Lalonde 2025-11-07 17:00:28 -05:00
  • 0f1c173d02 feat: bump versions to 1.4.1 (#3862) 1.4.1 Lorenze Jay 2025-11-07 11:19:07 -08:00
  • 19c5b9a35e fix: properly handle agent max iterations Greyson LaLonde 2025-11-07 13:54:11 -05:00
  • 1ed307b58c fix: route llm model syntax to litellm Greyson LaLonde 2025-11-07 13:34:15 -05:00
  • 5f1cfe22b9 Add comprehensive tests for MCP imports to address issue #3858 devin/1762530317-fix-issue-3858-mcp-imports Devin AI 2025-11-07 15:52:18 +00:00
  • d29867bbb6 chore: update version numbers to 1.4.0 1.4.0 Lorenze Jay 2025-11-06 20:04:44 -08:00
  • 81850350e8 Merge branch 'main' into gl/feat/pydantic-refactor-llms gl/feat/pydantic-refactor-llms Greyson Lalonde 2025-11-06 23:03:39 -05:00
  • b2c278ed22 refactor: improve MCP tool execution handling with concurrent futures (#3854) Lorenze Jay 2025-11-06 19:28:08 -08:00
  • f6aed9798b feat: allow non-ast plot routes Greyson LaLonde 2025-11-06 21:17:29 -05:00
  • 40a2d387a1 fix: keep stopwords updated Greyson LaLonde 2025-11-06 21:10:25 -05:00
  • 6f36d7003b Lorenze/feat mcp first class support (#3850) Lorenze Jay 2025-11-06 17:45:16 -08:00
  • fed3196ffe replace print by logger lg-debug-platform-studio Lucas Gomide 2025-11-06 20:45:00 -03:00
  • 7404d8f198 feat: restructure llms to pydantic Greyson Lalonde 2025-11-06 18:40:28 -05:00
  • 138b9af274 Merge branch 'main' into gl/feat/pydantic-refactor-llms Greyson Lalonde 2025-11-06 17:37:41 -05:00
  • 2997e6c3ad add debuging to Tool calls Lucas Gomide 2025-11-06 17:47:32 -03:00
  • 9e5906c52f feat: add pydantic validation dunder to BaseInterceptor Greyson LaLonde 2025-11-06 15:27:07 -05:00
  • fc521839e4 Lorenze/fix duplicating doc ids for knowledge (#3840) Lorenze Jay 2025-11-06 10:59:52 -08:00
  • a5e0803f20 Merge branch 'gl/feat/add-schema-validation-to-interceptor' into gl/feat/pydantic-refactor-llms Greyson Lalonde 2025-11-06 11:17:58 -05:00
  • c4279b0339 feat: add pydantic validation dunder to BaseInterceptor, improve HTTPTransport typing Greyson Lalonde 2025-11-06 10:42:31 -05:00
  • 965aa48ea1 feat: add pydantic validation dunder to BaseInterceptor, improve HTTPTransport typing Greyson Lalonde 2025-11-06 10:42:31 -05:00
  • 8105e74267 Fix agent loop not stopping after max iterations reached devin/1762441649-fix-max-iterations-loop Devin AI 2025-11-06 15:16:12 +00:00
  • 04b88bcf88 Add message_id to LLM events to differentiate stream chunks devin/1762437230-add-message-id-to-llm-stream-events Devin AI 2025-11-06 14:02:39 +00:00
  • 7a0feb8c43 Fix tool output token overflow issue devin/1762433678-fix-tool-output-token-overflow Devin AI 2025-11-06 13:01:52 +00:00
  • e4cc9a664c fix: handle unpickleable values in flow state gl/fix/align-llm-stop-sequences Greyson LaLonde 2025-11-06 01:29:21 -05:00
  • 7e6171d5bc fix: ensure lite agents course-correct on validation errors Greyson LaLonde 2025-11-05 19:02:11 -05:00
  • 4fd6863a02 chore: align json schemas with providers gl/fix/output-parser-exception-retry-logic Greyson Lalonde 2025-11-05 14:07:41 -05:00
  • 6111bb6c65 chore: remove pydantic schema parser Greyson Lalonde 2025-11-05 13:33:05 -05:00
  • fcb9941528 Fix stop sequences synchronization for Anthropic, Bedrock, and Gemini providers devin/1762366006-fix-stop-sequences-sync Devin AI 2025-11-05 18:15:31 +00:00
  • ae006fe0ad Merge branch 'main' into gl/fix/output-parser-exception-retry-logic Greyson Lalonde 2025-11-05 11:58:49 -05:00
  • 07ac8fb088 fix: run last llm call as structured output if passed Greyson Lalonde 2025-11-05 07:17:23 -05:00
  • 61ad1fb112 feat: add support for llm message interceptor hooks Greyson LaLonde 2025-11-05 11:38:44 -05:00
  • 54710a8711 fix: hash callback args correctly to ensure caching works Greyson LaLonde 2025-11-05 07:19:09 -05:00
  • 7380f7b794 feat: pass response model to lite agent Greyson Lalonde 2025-11-05 07:13:03 -05:00
  • de2e516995 fix: set hard stop to avoid infinite loop in parsing errors Greyson Lalonde 2025-11-04 21:17:41 -05:00
  • b45cb7a93e fix: set max reasoning attempts to address infinite loop Greyson LaLonde 2025-11-04 20:52:27 -05:00
  • b221db6061 chore: use json openapi schema instead of custom python schema Greyson LaLonde 2025-11-04 15:21:40 -05:00
  • 01a2e218ef feat: make prompt more robust Greyson LaLonde 2025-11-04 14:55:44 -05:00
  • 76f9d505ad fix: re-raise OutputParserError in format_answer for retry logic Greyson LaLonde 2025-11-04 14:37:29 -05:00
  • 97b573a3a4 Merge branch 'main' into gl/chore/deprecate-pydantic-schema-parser gl/chore/deprecate-pydantic-schema-parser Greyson LaLonde 2025-11-04 14:08:45 -05:00
  • 5abf976373 fix: allow adding RAG source content from valid URLs (#3831) Lucas Gomide 2025-11-04 07:58:40 -05:00
  • 8c0da6c129 Add regression tests for issue #3828: Flow with unpickleable objects in state devin/1762251954-fix-flow-deepcopy-rlock-v2 Devin AI 2025-11-04 10:27:39 +00:00
  • b58bbb7d83 Add regression tests for issue #3828: Flow with unpickleable objects in state devin/1762251449-fix-flow-deepcopy-rlock Devin AI 2025-11-04 10:23:00 +00:00
  • 010723ee83 tests: update test for new expectation Greyson Lalonde 2025-11-03 11:19:21 -05:00
  • 2e2eb77d2d Merge branch 'gl/chore/deprecate-pydantic-schema-parser' of https://github.com/crewAIInc/crewAI into gl/chore/deprecate-pydantic-schema-parser Greyson Lalonde 2025-11-03 11:05:10 -05:00
  • aed7788d8f chore: deprecate PydanticSchemaParser in favor of openapi spec schema Greyson Lalonde 2025-11-03 11:02:56 -05:00
  • d4c27312ee chore: deprecate PydanticSchemaParser in favor of openapi spec schema Greyson Lalonde 2025-11-03 11:02:56 -05:00
  • 329567153b fix: make plot node selection smoother Greyson LaLonde 2025-11-03 07:49:31 -05:00
  • 60332e0b19 feat: cache i18n prompts for efficient use Greyson LaLonde 2025-11-03 07:39:05 -05:00
  • 40932af3fa feat: bump versions to 1.3.0 (#3820) 1.3.0 Lorenze Jay 2025-10-31 18:54:02 -07:00
  • e134e5305b Gl/feat/a2a refactor (#3793) Greyson LaLonde 2025-11-01 02:42:03 +01:00
  • e229ef4e19 refactor: improve flow handling, typing, and logging; update UI and tests Greyson LaLonde 2025-11-01 02:15:06 +01:00
  • 27c08e48a7 chore: Regenerate uv.lock file devin/1761882750-fix-llm-config-in-crew-templates Devin AI 2025-10-31 03:57:48 +00:00
  • 888a1a9c77 Fix: Add LLM configuration examples and .env.example to crew templates Devin AI 2025-10-31 03:56:16 +00:00
  • 312c6d73bc Fix lint errors: Remove unnecessary assignments before return statements devin/1761822266-fix-additional-drop-params Devin AI 2025-10-30 11:10:22 +00:00
  • 26906113fe Fix: Apply additional_drop_params filtering across all LLM providers Devin AI 2025-10-30 11:04:53 +00:00
  • 5dc40bff9e Fix CrewDoclingSource filepath metadata misalignment devin/1761814608-add-metadata-to-knowledge-sources Devin AI 2025-10-30 09:30:37 +00:00
  • 58c5a2670a Fix whitespace in docstring for lint compliance Devin AI 2025-10-30 09:19:51 +00:00
  • 82bb304147 Add metadata support to Knowledge Source classes Devin AI 2025-10-30 09:04:35 +00:00
  • 2e9eb8c32d fix: refactor use_stop_words to property, add check for stop words Greyson LaLonde 2025-10-29 19:14:01 +01:00
  • 4ebb5114ed Fix Firecrawl tools & adding tests (#3810) Lucas Gomide 2025-10-29 13:37:57 -04:00