Pydantic Skills

Logfire and Pydantic AI plugins for Claude Code, Codex, and Cursor. Instrument your code with Logfire, query and export the resulting telemetry, and build LLM agents with Pydantic AI.

Plugins

Plugin	Hosts	Description	Capabilities
logfire	Claude Code, Codex, Cursor	Add Logfire observability and query/debug telemetry	Claude commands /instrument, /debug, /query; Codex/Cursor skills and MCP tools
logfire-exporter	Codex	Export Codex activity traces to Logfire	Codex lifecycle hooks
ai	Claude Code	Build AI agents with Pydantic AI	Pydantic AI skill docs
pydantic-ai-harness	Claude Code	Extend Pydantic AI agents with harness capabilities (Code Mode)	Pydantic AI Harness skill docs

Install In Claude Code

Add this marketplace to Claude Code:

claude plugin marketplace add pydantic/skills

Then install a plugin:

claude plugin install logfire@pydantic-skills
claude plugin install ai@pydantic-skills
claude plugin install pydantic-ai-harness@pydantic-skills

Install In Codex

Add the published Pydantic marketplace to Codex:

codex plugin marketplace add pydantic/skills

Then open Codex's plugin UI and enable the plugins you want from the Pydantic marketplace:

• Logfire - installs Logfire skills and the hosted Logfire MCP server.
• Logfire Exporter - installs Codex lifecycle hooks that export completed Codex turns and tool calls to Logfire.

Configure Logfire Exporter with a Logfire write token in your environment or ${XDG_CONFIG_HOME:-~/.config}/logfire-exporter/config.env. Restart Codex after configuration, and run /hooks if Codex asks you to review or trust the new hooks.

To use the EU Logfire MCP endpoint in Codex without editing plugin files, replace the MCP entry and re-authenticate:

codex mcp remove logfire
codex mcp add logfire --url https://logfire-eu.pydantic.dev/mcp
codex mcp login logfire
codex mcp get logfire

Start a new Codex conversation after switching so the MCP tools reload.

Troubleshooting Codex OAuth

Codex CLI 0.142.0 through at least 0.144.3 has an OAuth regression (openai/codex#31573): it drops the RFC 9207 iss parameter from the authorization callback, so codex mcp login can fail with Authorization server response missing required issuer even after the browser shows "Authentication complete". A server-side workaround is being rolled out to the hosted Logfire MCP servers (mid-July 2026). If you still hit this error (for example against a self-hosted Logfire instance that hasn't updated yet), either:

• downgrade Codex: npm install -g @openai/codex@0.141.0, or
• wait for a Codex release that fixes the issue, then retry codex mcp login logfire.

Install In Cursor

Install the Logfire plugin from the published pydantic/skills repository:

git clone https://github.com/pydantic/skills.git
mkdir -p ~/.cursor/plugins/local
cp -R skills/plugins/logfire ~/.cursor/plugins/local/logfire

Then restart Cursor or run Developer: Reload Window. The plugin configures:

• display name Logfire
• Logfire skills for instrumentation, querying, and UI-opening workflows
• hosted Logfire MCP server from mcp.json

The Logfire MCP server requires normal Logfire authentication, such as logfire auth or a suitable LOGFIRE_TOKEN.

Cross-Agent Skills

The skills/ directory contains standalone SKILL.md files compatible with 30+ agents via the agentskills.io standard - including Codex, Cursor, Gemini CLI, and Claude Code.

Skill	Description
logfire-instrumentation	Add Logfire observability to Python, JS/TS, and Rust apps
logfire-query	Query and analyze Logfire traces, logs, spans, metrics, and activity data
logfire-ui	Open Logfire project pages, live views, traces, and Explore filters
building-pydantic-ai-agents	Build LLM-powered agents with Pydantic AI — tools, capabilities, streaming, testing
pydantic-ai-harness	Extend Pydantic AI agents with harness capabilities like Code Mode (sandboxed run_code)