CrewAI

持续追踪核心框架、最新模型规格与重要系统接口的更新迭代轨迹。

1.15.44

1.15.44

What's Changed Features Add capability to save files as versioned, shareable artifacts. Bug Fixes Fix inline agent and crew action references relative to the flow file. Stop duplicating kept separators when merging chunks. Isolate nested defaults from a shared flow declaration. Restrict Pillow to PNG, JPEG, GIF, and WEBP when opening images. Reset cache eviction order on clear. Treat a docs path on GitHub as a repository. Exit with the script's status when crewai train fails. Relax av pin to ensure current FFmpeg installs. Keep a deeply nested span that the run app cannot render as text in one...

1.15.43

1.15.43

What's Changed Bug Fixes Fix detection of multimodal support for Claude 5. Documentation Update snapshot and changelog for v1.15.42. Contributors Vidit Ostwal (@Vidit-Ostwal) , Anubhav Lal (@anubhavlal07) , João Moura (@joaomdmoura)

1.15.42

1.15.42

What's Changed Features Make ChromaDB an optional extra Default Knowledge to local Qdrant and decouple storage from ChromaDB Pick tools for standalone agents and add them later Bug Fixes Store output_json values as JSON to prevent date issues in .json and output_file Give each Flow its own ID from an initial_state instance Honor max_execution_time without waiting for the worker Save task outputs that hold URLs and other Pydantic types Read the experiment baseline as UTF-8 Prefix sanitized tool names that start with a digit Count scoped LanceDB records exactly above 50,000 rows Report Serply's ...

1.15.41

1.15.41

What's Changed Features Allow approved private URL hosts. Bug Fixes Fix projects to depend on PyMuPDF 2.x. Drop a real storage URL from the URL read tests. Keep earlier step outputs across a human-feedback resume. Strip cache_breakpoint marker on the LiteLLM path. Documentation Snapshot and changelog for v1.15.40. Contributors Vidit Ostwal (@Vidit-Ostwal) , Gabe Milani (@gabemilani) , @github-actions[bot], Lorenze Jay (@lorenzejay)

1.15.40

1.15.40

What's Changed Features Remove chromadb from the embedding function layer. Add exact historical state lookup by row id. Wire jsonl into rag data type discovery. Scaffold standalone agents as JSON by default. Add backend record lifecycle and pagination tests. Bug Fixes Run after_llm_call hooks on async calls. Stop an agent whose apps CrewAI AMP cannot resolve. Documentation Update Arize agent evaluation guide URL. Contributors Aadesh Salecha (@AadeshSalecha) , Kiko0301 , Vidit Ostwal (@Vidit-Ostwal) , Dat Daryl Ngo (@arizedatngo) , Brooke Ballhaus (@bballhaus) , Amy Chang (@chang-amy) , factory...

1.15.39

1.15.39

What's Changed Documentation Document the apps_not_connected error on /kickoff and /status Contributors João Moura (@joaomdmoura)

1.15.38

1.15.38

What's Changed Features Make gpt-6-luna crewAI's default model everywhere Read spreadsheet URLs in an agent's input with tools over every row Tag eval usage spans with target and a per-evaluation id Add --no-open to crewai eval for headless use Bug Fixes Count a crewai eval run that exits with an error as run_failed Create the project .env owner-only when crewai eval turns tracing on Upload a failed anonymous run that crewai eval started Check openai-compatible providers' API key on the first call Scope memoized crew methods to their instance Exit non-zero when crewai run's project script fail...

[ 重大突破 ]1.15.33

1.15.33

What's Changed Features Add crewai skill search to list the organization's skills Pass crewai run --inputs to a Python flow Add deploy create --zip and print the commit a zip is built from Bug Fixes Report token usage for LiteLLM tool-call and structured-output calls Restore recent backwards-compatible imports and token Let the JSON crew runner load against an older crewai-core Keep deeply nested JSON as text in the run TUI on Python 3.12 Export persist from crewai.flow.flow Load JSON crews in crewai chat Treat an env tracing false as off in is_tracing_enabled Exit with guidance when a Python ...

[ 重大突破 ]1.15.34

1.15.34

What's Changed Features Add standalone LiteLLM catalog PR before version bump Record where crewai run stops before kickoff Add crewai skill search to list the organization's skills Pass crewai run --inputs to a Python flow Add deploy create --zip and print the commit a zip is built from Document agent planning with planningconfig and migrate reasoning page Bug Fixes Stop offering shut-down Gemini models in crewai create Raise MissingAPIKeyError for OpenAI-compatible providers Count repeated tool call stops in anonymous telemetry Count a crewai run whose project script fails Let crewai[litellm]...

1.15.35

1.15.35

What's Changed Documentation Update litellm model catalog for v1.15.35 Add snapshot and changelog for v1.15.34 Contributors @github-actions[bot], João Moura (@joaomdmoura) , Lorenze Jay (@lorenzejay)

1.15.36

1.15.36

What's Changed Features Add experimental conversational Agent sessions. Ask for a missing API key when crewai run starts. Record the error class when crewai run cannot load a project. Name the apps CrewAI AMP cannot resolve for an agent. Bug Fixes Fix async call execution in the caller's context for sync tools. Resolve issue with empty value precedence in the run's key check. Stop a JSON crew run cleanly when its API key is missing. Call GPT-6 Astra and GPT-6.1 Sol tools through the Responses API. Handle the sourcedAnswer output shape correctly. Drop None metadata values before writing to Chro...

1.15.37

1.15.37

What's Changed Features Add functionality to read spreadsheets for agents in crewai-files. Scaffold and load standalone agent projects. Documentation Update RagTool description for better parsing by Mintlify. Contributors Vidit Ostwal (@Vidit-Ostwal) , João Moura (@joaomdmoura) , Lorenze Jay (@lorenzejay)

1.15.31

1.15.31

What's Changed Features Support conversational flow context compaction. Bug Fixes Update vulnerable audit dependencies. Stop sending 'stop' to GPT-6 and newer models. Run the provider listing off the loop in acleanup_provider_files . Emit LLMGuardrailCompletedEvent for a malformed guardrail result. Name the renamed integration token when it is missing. Fix path traversal in output_file via template interpolation. Patch deserialization, SQL injection, and path traversal vulnerabilities. Restore configured logging levels after suppression. Drop checkpoint resume state when copying a crew. Honor ...

1.15.32

1.15.32

What's Changed Bug Fixes Fix failing flaky tests Documentation Update snapshot and changelog for v1.15.31 Contributors @lorenzejay

1.15.29

1.15.29

What's Changed Features Add support for LiteLLM-compatible kwargs. Implement platform application search in CLI. Highlight action picker search in CLI. Trace conversation routes and public replies in flow. Select organization during login in CLI. Add offline jev routing and guardrail evaluation runner. Run integration app actions directly with call: tool. Promote AgentExecutor out of experimental status. Add Azure DocumentDB vector search integration. Add read-only notebook text loader and standard-library jsonl loader. Ensure crewAI eval keeps the evaluation's full data in the project. Enable...

1.15.30

1.15.30

What's Changed Bug Fixes Fix AGENTS.md to start the login process while evaluating and stop after three rounds. Documentation Update snapshot and changelog for v1.15.29. Contributors @joaomdmoura

1.15.28

1.15.28

What's Changed Features Add crewai eval command to report where a stopped model broke. Use LiteLLM catalog for context windows. Bug Fixes Fix flow to learn from resumed human feedback and preserve reviewed output types. Warn when non-empty apps resolve to zero platform tools. Fix flow add-crew to correctly find the package from pyproject. Show reasons for deployment failures and fix deploy remove and lockfile checks. Fix trigger list and payload requests timeout to 60 seconds. Read AMP's source_type during deploy push to ensure ZIP deployments use the new code. Decode local JSON files with UTF...

[ 重大突破 ]1.15.27

1.15.27

What's Changed Features Add XPU to OpenCLIP provider device options Add deepinfra as an OpenAI-compatible provider Record why an evaluation stopped before its verdict Record what the run app's Deploy met before an attempt Implement per-run cost and time tracking in crewai eval --models without repeated progress lines Bug Fixes Fix issue where stopSequences were sent to OpenAI GPT-6/GPT-5.6/gpt-oss Preserve GitHub loader source attribution Fix SQLite flow initialization contention Documentation Add snapshot and changelog for v1.15.26 Contributors @StoyanD , @ats3v , @joaomdmoura , @kimnamu , @l...

1.15.26

1.15.26

What's Changed Bug Fixes Fix ownership retention of rwlock when acquisition is interrupted. Resolve issue with documentation site source URLs not being preserved. Fix task output display in log-tasks-outputs to show actual output. Documentation Update snapshot and changelog for version v1.15.25. Contributors @Velmet44 , @Vidit-Ostwal , @joaomdmoura , @leilei3167 , @lorenzejay , @lux-liang

1.15.25

1.15.25

What's Changed Documentation Snapshot and changelog for v1.15.24 Contributors @lorenzejay

1.15.24

1.15.24

What's Changed Features Add crewai eval to print a markdown brief when run by an agent Queue relevant background replies Add experimental turn and reply identities Add Oracle Integrations Add experimental job lifecycle and runner Add crewai eval --models, and llm_overlay to swap models as well as roles Bug Fixes Fix train filename error message in CLI Allow a start step to re-run on the events it listens to Let human_feedback emit steps expose the review in outputs Ensure crewai eval exits with code 1 unless the gate passed and prompt to log in when nothing was traced unattended Report tracing...

1.15.23

1.15.23

What's Changed Features Add support for native Gemini 3.8 Flash Implement evaluation of the last traced run through AMP in crewai eval Record the last traced run for crewai eval instead of printing it Improve platform integration setup UX Prioritize popular platform integrations Enhance tracing task spans to include declared output format and results Bug Fixes Fix visibility of the panel for viewing traces Resolve issue with turning tracing on and add an Evaluate button in the TUI Retry throttled provider calls in LLM Fall back to synchronous calls from acall in Bedrock Print areas graded by t...

[ 重大突破 ]1.15.22

1.15.22

What's Changed Features Support aliases as connection identifiers Record reasons for deployment creation failures Collect human feedback and pause events in tracing Add llm_overlay context variable to route agent roles to models Carry task_prompt and output in agent execution payloads Validate platform integrations during crew setup Add platform tools to JSON crew wizard Expose CrewAI Platform application catalog Add OpenRouter as a supported embedding provider Bug Fixes Accept CRLF in inline skill definitions Load text file URLs through the safe fetcher Close SQLite connections in kickoff tas...

[ 重大突破 ]1.15.21

1.15.21

What's Changed Features Add telemetry to track checkpoint runtime and CLI usage. Bug Fixes Fix gateway errors reported inside an HTTP 200 response. Keep deploy push on the AMP create source. Report scrape failures instead of raising IndexError in the Oxylabs integration. Route all DashScope models through the native provider. Persist JSON checkpoints as UTF-8. Drop stray $ in f-string search URLs for BrightData. Support list-form type arrays in JSON schema conversion. Correct gpt-4o-mini context window from 200000 to 128000. Preserve streaming tool call arguments at contentBlockStop. Make pre-...

1.15.19

1.15.19

What's Changed Features Add Clipper integrations client Add now() to the CEL expression environment Record how a crew run ended for every user Report machine size as a coarse band, not a core count Add injectable client for CrewAI platform tools Bug Fixes Fix reading of octet-stream and xlsx URLs in urlreadtool Fix appending trailing user turn in native Gemini provider Normalize scheme and port in Ollama base URL Preserve reusable scope configs in memory Bump pypdf to 6.16.2 for security vulnerability Bump nltk to 3.10.3 for security vulnerability Fix native structured outputs for current Clau...

1.15.20

1.15.20

What's Changed Bug Fixes Fix legacy platform tool alias discovery Documentation Update snapshot and changelog for v1.15.19 Contributors @joaomdmoura , @vinibrsl

1.15.17

1.15.17

What's Changed Features Add declarative conversational flows documentation Synthesize built-in conversational methods for declarations Enable declarations to drive conversational mode Make conversational opt-in unmistakable Carry the AMP slug on tools resolved from a slug reference Handle oversized single messages during chunking Bug Fixes Fix usage of the URL hostname as MCP HTTP and SSE server_name Close the agent scope on every failed attempt Attribute tool errors to the tool that failed Pin SSRF checks to each redirect hop and peer IP Resolve issues with native tool calls broken over OpenA...

1.15.18

1.15.18

What's Changed Features Promote conversational flows to stable Record a created deployment with the given UUID Enhance conversational flow documentation and APIs Let a declaration name the router's response format Let a chat flow declare its own state shape Accept crew-style LLM config in a conversational declaration Report project creation with the minted ID Record whether a run had inputs, without recording the inputs Backfill project ID from every user-invoked project command Bug Fixes Preserve tool results when the final answer is empty Map default Claude Sonnet 4.6 to its 1M context windo...

1.15.16

1.15.16

What's Changed Features Introduce execution context management with UUID support Record what kind of exception ended a flow Record when a trace batch is shared with AMP Count deployments from any origin and record where they started Bug Fixes Record the running release on every emitted span Fix MySQL search table name validation Stop a failed turn from marking the next one as failed Documentation Add Frontend guides for CopilotKit and AG-UI Contributors @joaomdmoura , @lorenzejay , @ranst91 , @theCyberTech

1.15.15

1.15.15

What's Changed Features Report flow outcome, duration, and human-in-the-loop signals. Bug Fixes Emit FlowStartedEvent when a boundary hook aborts the flow. Scope span export to our own tracer provider. Bump torch to version 2.13.0 to address security vulnerability. Bump gitpython to version 3.1.58 in crewai-tools[github]. Refactoring Update date injection functionality in agents. Standardize CLI flags to kebab-case. Documentation Snapshot and changelog for v1.15.14. Contributors @Vidit-Ostwal , @joaomdmoura , @lorenzejay , @lucasgomide , @theCyberTech

1.15.14

1.15.14

What's Changed Features Split runtime context from coding agent and add project ID Documentation Update snapshot and changelog for v1.15.13 Contributors @joaomdmoura

[ 重大突破 ]1.15.13

1.15.13

What's Changed Bug Fixes Fix preservation of provider on LiteLLM-routed models. Harden brittle LLM event-bus mocks. Fix underreporting of Anthropic cache token usage. Bump h2 to version 4.4.1 to address security vulnerability GHSA-6hr6-w5qg-qmwg . Documentation Add DOCS_TRANSLATIONS workflow for locale synchronization. Fix broken README links, table of contents, and contribution guidance. Snapshot and changelog for version 1.15.12. Contributors @Vidit-Ostwal , @joaomdmoura , @lucasgomide , @theCyberTech

1.15.12

1.15.12

What's Changed Features Bump Flow canary on release Add URLReadTool for reading arbitrary URLs Add app metadata to platform action tools Unify scaffolding under crewai create <resource> Bug Fixes Clarify conversational route/handler name collision errors Documentation Update scaffold AGENTS.md for unified create CLI Breaking Changes None Contributors @Vidit-Ostwal , @joaomdmoura , @lorenzejay , @lucasgomide , @theCyberTech

1.15.11

1.15.11

What's Changed Features Track interception-hook dispatches in telemetry Add project_id to link OSS usage to an enterprise account Surface AMP in AGENTS.md and detect coding agents in telemetry Add IBM Db2 search tool Bug Fixes Clear CodeQL incomplete URL substring sanitization alerts Bump aiohttp and cryptography to clear six GHSA advisories Report the real CEL error for failures inside map literals Correctly skip code CI for docs-only PRs Documentation Snapshot and changelog for v1.15.10 Contributors @PawanThakurIBM , @Vidit-Ostwal , @gabemilani , @joaomdmoura , @lorenzejay , @lucasgomide , @...

1.15.10

1.15.10

What's Changed Features Collect skill usage events Documentation Remove migrated AMP documentation Update security reporting guidelines Snapshot and changelog for v1.15.9 Contributors @gvieira , @joaomdmoura , @theCyberTech

1.15.9

1.15.9

What's Changed Features Surface tool failures instead of reporting them as success Emit FlowFailedEvent when a flow execution fails Implement progressive disclosure for skills Documentation Update snapshot and changelog for v1.15.8 Contributors @github-actions[bot], @joaomdmoura , @lorenzejay , @lucasgomide

1.15.8

1.15.8

What's Changed Features Add WaitTool for pausing on long-running jobs. Bug Fixes Fix FileWriterTool writes and address rough edges in file tool. Mark E2B_API_KEY as a required env var for E2B tools. Documentation Refresh model availability guidance. Contributors @github-actions[bot], @joaomdmoura , @lucasgomide , @oalami , @thiagomoretto

1.15.6

1.15.6

What's Changed Bug Fixes Fix detection of Anthropic preview tool-use blocks. Preserve strict tool schema property names. Dispatch execution_end hook on failed crew and flow executions. Handle async get_agent in load_agent_from_repository. Fix dependency resolution issues. Documentation Snapshot and changelog for v1.15.5. Contributors @alex-clawd , @iris-clawd , @lorenzejay , @lucasgomide , @theCyberTech , @vinibrsl

[ 重大突破 ]1.15.7

1.15.7a1

What's Changed Bug Fixes Fix registry skills resolution through the runtime's CrewAI+ client. Recover from the GPT-5.6 tools and reasoning_effort 400 errors. Make tool calling work on the Responses API path. Route responses-only models to prevent 404 errors. Bump bedrock-agentcore dependency to patch CVE-2026-16796 . Observability Emit skill usage events at runtime for improved observability. Documentation Snapshot and changelog updates for version 1.15.6. Contributors @alex-clawd , @joaomdmoura , @lorenzejay

[ 重大突破 ]1.15.7

1.15.7

What's Changed Bug Fixes Resolve registry skills through the runtime's CrewAI+ client Recover from the GPT-5.6 tools + reasoning_effort 400 Make tool calling work on the Responses API path Route responses-only models instead of failing with 404 Bump bedrock-agentcore to patch CVE-2026-16796 Observability Emit skill usage events at runtime for observability Documentation Add snapshot and changelog for v1.15.7a1 Contributors @alex-clawd , @joaomdmoura , @lorenzejay

1.15.5

1.15.5

What's Changed Features Authenticate skill registry downloads Documentation Update snapshot and changelog for v1.15.4 Contributors @vinibrsl

1.15.3

1.15.3a1

What's Changed Features Add organization ID parameter to PlusAPI client. Add step interception points and rework execution hooks documentation around @on . Wire execution-boundary interception points. Add generic interception-hook dispatcher. Run declarative flows on the TUI (headless terminal fallback). Improve custom OpenAI URLs. Bug Fixes Fix null repository agent attributes. Fix after_llm_call hooks to prevent breaking native tool execution. Stop double-appending the turn reply when a handler trims history. Make tool-result caching opt-in instead of on by default. Stop rewriting the author...

1.15.3

1.15.3a2

What's Changed Bug Fixes Fix synchronization of kickoff-completed event with OUTPUT hook result Documentation Update snapshot and changelog for v1.15.3a1 Dependency Updates Bump setuptools to 0.83.0 to address PYSEC-2026-3447 Contributors @lucasgomide , @vinibrsl

1.15.3

1.15.3

What's Changed Features Add organization ID parameter to PlusAPI client Add step interception points and rework execution hooks documentation around @on Wire execution-boundary interception points Add generic interception-hook dispatcher Run declarative flows on the TUI (headless terminal fallback) Bug Fixes Sync kickoff-completed event with OUTPUT hook result Fix null repository agent attributes Ensure after_llm_call hooks do not break native tool execution Avoid double-append of the turn reply when a handler trims history Make tool-result caching opt-in instead of on by default Stop rewritin...

1.15.4

1.15.4

What's Changed Features Promote Skills Repository out of experimental status Documentation Add Flows in Studio documentation Contributors @jessemiller , @joaomdmoura , @vinibrsl

1.15.2

1.15.2

What's Changed Features Pull latest LLM models dynamically in the crew wizard. Support inline skill definitions. Add generated Flow Definition authoring skill. Support templated Flow action inputs. Add text helper for flow CEL prompts. Add text helper to flow skill example. Implement message setup and feedback handling in AgentExecutor. Add repository agents to flow definitions. Define stream frame protocol for flows. Type tool and app in CrewDefinition. Repoint template commands to crewAIInc-fde org. Bug Fixes Key model-catalog cache by exact API key, shorten TTL, and skip Ollama. Unify crewa...

1.15.2

1.15.2a2

What's Changed Features Add aiobotocore to the bedrock extra Document flow agent options Add text helper to flow skill example Add text helper for flow CEL prompts Add streaming docs to the navigation Bug Fixes Reject self-listening flow methods Documentation Update snapshot and changelog for v1.15.2a1 Squeeze AGENTS.md file Contributors @akaKuruma , @github-code-quality[bot], @lorenzejay , @vinibrsl

1.15.2

1.15.2a1

What's Changed Features Repoint template commands to crewAIInc-fde org Support inline skill definitions Define stream frame protocol for flows Add type tool and app in CrewDefinition Add generated Flow Definition authoring skill Bug Fixes Cut docs version navigation from Edge to prevent new pages from being dropped Documentation Document Cost Limit rule type in Agent Control Plane Drop CREWAI_LOG_FORMAT references from Datadog guide Contributors @danielfsbarreto , @joaomdmoura , @lorenzejay , @lucasgomide , @vinibrsl

1.15.1

1.15.1

What's Changed Features Initialize Git repositories for generated projects ( #6364 ) Require explicit CrewAI project definitions ( #6358 ) Open deployment page after CLI deploy ( #6343 ) Bug Fixes Fix deployment page link ID resolution ( #6365 ) Fix JSON crew template rendering ( #6359 ) Fix JSON crew version pin ( #6342 ) Fix SSRF redirect bypass in scraping fetches ( #6331 ) Documentation Improve open source positioning in README ( #6363 ) Improve coding agent setup call-to-action ( #6344 ) Add snapshot and changelog for version 1.15.1a1 ( #6362 ) Contributors @joaomdmoura , @lorenzejay , @o...

1.14.8

1.14.8a4

What's Changed Features Support conversational flows in the CLI TUI. Bug Fixes Fix symlink path traversal in skill archive extraction. Validate declarative flow definition paths. Documentation Update snapshot and changelog for v1.14.8a3. Contributors @lorenzejay , @theCyberTech , @vinibrsl