{"full_name":"crewAIInc/crewAI","owner":"crewAIInc","name":"crewAI","shard":"1","url":"https://github.com/crewAIInc/crewAI","description":"Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.","homepage":"https://crewai.com","topics":["agents","ai","ai-agents","llms","aiagentframework"],"tag_pages":["agents","ai","ai-agents","llms"],"language":"Python","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/170677839?v=4","created_at":"2023-10-27T03:26:59Z","pushed_at":"2026-11-15T19:27:48Z","metrics":{"stars":60954,"forks":9197,"watchers":390,"issues":2934,"prs":6012,"releases":255,"issues_open":498,"issues_closed":2436,"prs_open":776,"prs_closed":2662,"prs_merged":2574,"discussions":0},"growth":{"day":{"stars":58,"forks":13,"base_day":"2026-11-15"},"week":{"stars":275,"forks":80,"base_day":"2026-11-09"}},"history":{"dates":["2026-11-09","2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[60679,60720,60769,60809,60823,60861,60896,60954],"forks":[9117,9136,9148,9160,9161,9172,9184,9197],"watchers":[386,386,387,387,389,389,389,390],"issues_open":[null,null,496,495,500,495,501,498],"prs_open":[null,null,752,765,766,767,772,776]},"boards":{},"same_owner":[],"issues":[{"number":5802,"title":"Tool re-execution on task retry has no idempotency guard — duplicate payments, emails, trades possible","comments":225,"created_at":"2026-05-14T03:00:55Z"},{"number":5888,"title":"[FEATURE]:Governance middleware hook for tool call authorization","comments":160,"created_at":"2026-05-21T11:37:57Z"},{"number":6025,"title":"[FEATURE] Runtime release-control mediation layer before agent/tool execution","comments":127,"created_at":"2026-06-03T07:53:04Z"},{"number":6414,"title":"[FEATURE] Native deterministic guardrail to prevent infinite agent delegation and tool loops","comments":54,"created_at":"2026-07-01T14:39:09Z"},{"number":6043,"title":"[FEATURE] Memory write guards for multi-agent crews — prevent cross-agent memory poisoning","comments":31,"created_at":"2026-06-04T20:37:16Z"},{"number":6221,"title":"Agent governance: deterministic tool permission gating for crew agents","comments":15,"created_at":"2026-06-18T17:02:06Z"},{"number":6072,"title":"[BUG] `human_input=True`: the feedback prompt references a \"Final Result above\" that is never displayed unless `verbose=True`","comments":14,"created_at":"2026-06-08T20:12:44Z"},{"number":8932,"title":"[BUG] MCP tool calls are sent again after a timeout, so one action can run a tool up to 3 times","comments":12,"created_at":"2026-11-02T05:48:20Z"}],"discussions":[],"latest_release":{"name":"1.15.41","tag":"1.15.41","published_at":"2026-11-12T18:42:07Z","body":"## What's Changed\n\n### Features\n- Allow approved private URL hosts.\n\n### Bug Fixes\n- Fix projects to depend on PyMuPDF 2.x.\n- Drop a real storage URL from the URL read tests.\n- Keep earlier step outputs across a human-feedback resume.\n- Strip cache_breakpoint marker on the LiteLLM path.\n\n### Documentation\n- Snapshot and changelog for v1.15.40.\n\n## Contributors\n\n@Vidit-Ostwal, @gabemilani, @github-actions[bot], @lorenzejay","is_prerelease":false},"readme":true,"readme_zh":false,"news":[],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}