Write pytest tests for AutoReview modules with mocked LLM responses.
npx skills add https://github.com/mcleanT/AutoReview --skill write-tests-mcleant
Please help me install this Agent Skill. Skill: write-tests Source: https://github.com/mcleanT/AutoReview/tree/main/.claude/skills/write-tests Command: npx skills add https://github.com/mcleanT/AutoReview --skill write-tests-mcleant
This Skill codifies AutoReview's testing conventions so contributors can write reliable, consistent pytest tests that correctly mock LLM behavior, handle asynchronous code, and reuse shared fixtures without guessing project patterns.
Write a pytest test file for a target module that defines a local MockLLM returning structured LLM responses and uses the provided conftest fixtures and make* builders.
High-intent search queries and answers about installing and using this skill.
To write pytest tests mocking LLM responses, set asyncio_mode to auto and define local Mock*LLM classes returning LLMResponse or LLMStructuredResponse for deterministic assertions. This ensures reliable testing of asynchronous functions and structured outputs.
The best way to structure test data builders is using module-level _make_* builders. This enforces consistent test data creation conventions and allows tests to reuse provided conftest fixtures for sample data across unit and integration tests.
Yes, you can patch external APIs in integration tests using provided conftest fixtures. Mark tests as integration when covering external API mocking, and use local Mock*LLM classes to wrap responses in LLMStructuredResponse for deterministic parsing assertions.
Mocking LLM outputs works by wrapping responses in LLMStructuredResponse objects returned from local Mock*LLM classes. This enables tests to assert structured parsing behavior, prompt content, and response model identity without live API calls.
Yes, asyncio_mode must be set to auto when testing asynchronous LLM functions. This configuration enables pytest to automatically handle async test functions, allowing reliable execution of tests covering asynchronous extraction and synthesis components.
Mark pytest tests as integration tests when they involve patching external APIs. Unit tests should use local Mock*LLM classes with LLMResponse wrappers, while integration tests cover broader interactions with external API mocking and shared conftest fixtures.
timequityGenerate pytest tests covering unit, integration, and async patterns.
vndkubiGenerate JUnit 5 tests with AssertJ assertions for Java classes.
ArabelaTsoGenerate Python integration tests with pytest fixtures and test data builders.
thechandanbhagatGenerate pytest and Jest unit tests, integration tests, and mock data.
