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
- `CodeMode` under Temporal: a snippet that computes for over 2 s blocks the workflow worker and leaves Monty running
- Persistent `shell` on `SSHWorkspace`: a command over 128 KiB is reported as a pending launch, and the run ends
- Document migrating to `ModelFallback` from a run-level fallback loop, plus latency-budget and auth-error recipes
- Anthropic `server_tool_use` filter ignores `extra_body['tools']`
- Docs: say in the Temporal guide that upgrading across 2.54.0 can break in-flight workflows
- Retry a tool call whose validation error input isn't JSON
- Gate `OpenAIEmbeddingModel.max_input_tokens`/`count_tokens` on an embedding profile, not `system != 'openai'`
- Keep a sub-agent's work when it reaches its own usage or time budget in `SubAgents`
- Make the compaction hooks Conductor relies on public before harness v1
- Add per-tool call budgets and repeat guards that answer the model instead of raising
- Docs
- Python not yet supported