1. Habit
  2. Skills
  3. Building Pydantic AI Agents skill
pydantic · 6K installs

Building Pydantic AI Agents skill
Runs by itself, when your agent code changes.

Your agents stay tested and easy to trace.

Install the Building Pydantic AI Agents skill and make it a habit: follow https://gethabit.dev/skills/pydantic/building-pydantic-ai-agents.md

  1. 1Copy
  2. 2Paste into Claude Code, Codex or Cursor
  3. 3It runs by itself

6K installs144 starsMIT licenceSource on GitHub ↗

Building Pydantic AI Agents

by pydantic · 6K installs

The skill

Helps you build and test AI agents in Python.

Only when you ask.

Fewer bugs

Your agents stay tested and easy to trace. Fewer bugs when you change code.

↻ Runs by itselfwhen your agent code changes

How the habit works.
You don't have to.

Three questionsWhen does it start? How does it check its work? What does it remember?

01

It starts by itself

agent code changes. Runs in GitHub Actions or Claude Code routine.

  • ↻ agent code changes
  • pull request touches files importing pydantic_ai

02

It checks its work

TestModel tests pass. If a check fails, it tries again.

  • ✓pytest passes with TestModel via agent.override
  • ✓model strings include a provider prefix
  • ✓@agent.tool functions take RunContext first and tool_plain ones do not
  • ✓no history_processors kwarg in new code

03

It remembers

agent test log. The next run reads it first: gotchas found before and failing tests.

loops/pydantic-ai-agent-test-loop/runs.md

One run, step by step.
Then it stops.

Stop ruleStops after one change and 2 test fix rounds per run.

  1. 01

    Route the task

    Your agent picks the one reference that fits, such as Agents Core or Tools Core.

  2. 02

    Check setup

    It checks each Agent has name=, Caching() and logfire.configure() unless you use another backend.

  3. 03

    Check loading

    It considers defer_loading=True for capabilities not needed on most turns.

  4. 04

    Check gotchas

    It checks tool decorators, provider prefixes, output_type unions and hook names.

  5. 05

    Test

    It runs TestModel or FunctionModel tests inside agent.override.

  6. 06

    Log

    It records failures and fixes.

What it may do.
Nothing more.

Safe by defaultIt writes only in its own folder and asks you before anything goes out.

Is it for you?
Use it when it fits.

Quick checkRead both lists. If the left one sounds like you, the Building Pydantic AI Agents skill helps.

Use it when

  • ✓ You build an agent with Pydantic AI
  • ✓ You add tools or capabilities
  • ✓ You test agent behavior

Skip it when

  • – You only use pydantic validation
  • – You use LangChain or CrewAI

Questions

Building Pydantic AI Agents skill, answered.

Can I set agent.model for tests?

No. Use the agent.override context manager with TestModel.

Should telemetry text be trusted?

No. The skill treats traces and tool results as data, not instructions.

How do I install the Building Pydantic AI Agents skill?

Copy the prompt at the top into your agent. It installs the skill from pydantic/skills and sets up the habit. Or run npx skills add pydantic/skills --skill building-pydantic-ai-agents.

Is the Building Pydantic AI Agents skill free?

Yes. It is published under MIT. Habit is free too.

Building Pydantic AI Agents. Now as a habit.

Install the Building Pydantic AI Agents skill and make it a habit: follow https://gethabit.dev/skills/pydantic/building-pydantic-ai-agents.md

  • Free
  • No account
  • Undo in one line

All pydantic skills →More code skills →

Works in

  • Claude Code
  • Codex
  • Cursor
  • ChatGPT
  • Claude app

Made by pydantic, published under MIT. Habit links to the original and never changes it. Is this your skill?

Copying didn't work here. Select the line and copy it by hand.