pydantic-ai
https://github.com/pydantic/pydantic-ai
Python
GenAI Agent Framework, the Pydantic way
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported6 Subscribers
Add a CodeTriage badge to pydantic-ai
Help out
- Issues
- AgentRunResult acknowledgement override targets a failed same-named output call
- Harness E2B sandbox: allow passing E2B network options (allow_public_traffic)
- Pre-execution shell guardrails for [pydantic-ai]
- Read HuggingFace native `reasoning` fields into thinking parts
- Accept full-precision score answers from System One decision models
- Feature request: Support offline evaluation of precomputed outputs without executing a task
- Map google-genai `IMAGE_RECITATION`, `TOO_MANY_TOOL_CALLS`, and `IMAGE_OTHER` finish reasons
- Approved tool runs again if the process dies after it executed but before the history was saved
- Let an explicit `output_type=str` override the spec's `output_schema` in `Agent.from_spec` and `Agent.from_file`
- Add a `CapabilityPart` base so capabilities stop hand-building tags, voices and escaping
- Docs
- Python not yet supported