Pydantic's hosted Logfire MCP server lets agents run SQL over your traces and logs, find exceptions by file, and manage dashboards, alerts, notification channels and issues. It has US and EU endpoints over Streamable HTTP with browser OAuth, or an API key with project:read scope as a bearer token. The older local server is deprecated.
Also available as
Connect
- 1Connect it in ChatGPT
Turn on Developer mode in ChatGPT (Settings → Security and login), open Plugins, select + and paste https://logfire-us.pydantic.dev/mcp as the MCP server URL, then sign in with OAuth. Add it to a chat from the composer's tools menu.
- 2Or add it to Codex or the API
codex mcp add logfire --url https://logfire-us.pydantic.dev/mcp codex mcp login logfireThe login command opens a browser to finish OAuth.
Configs are generated from the publisher's documented endpoint. Check their docs for plan requirements and scopes. New to this? Read the step-by-step tutorial.
Tools 9
query_runquery_schema_referencequery_find_exceptions_in_fileproject_listdashboard_createdashboard_add_panelalert_createalert_listissue_listThings to ask
- “What exceptions were raised in app/payments.py in the last hour?”
- “Show the slowest endpoints over the last 24 hours”
- “Create an alert when error rate exceeds 5%”
- “Build a dashboard of request latency by route”