Testing Standards

This skill defines the requirements and best practices for testing within the ai-context-core project. We prioritize environment consistency and execution speed.

1. Test Framework & Runner

• •Authoring Style: Tests are written using the standard unittest library (subclassing unittest.TestCase).
• •Execution Runner: We use pytest as the test runner to leverage its superior CLI reporting, fixture management, and coverage tools.
• •Location: All tests must reside in the tests/ directory.
• •Naming: Files must be prefixed with test_ (e.g., test_engine.py).

2. Docker & Environment Isolation

All tests must be verifiable in an isolated containerized environment to replicate CI conditions.

• •Mandatory: Use the project's Makefile to run tests via Docker.
  bash

  make docker-test
  

• •Local Dev: You may run uv run pytest for quick feedback loops, but final validation requires Docker.

3. Coverage Requirements

• •Minimum Coverage: The project maintains a threshold of 70% code coverage.
• •Critical Paths: Core logic in engine.py and ast_utils.py should aim for >80%.

4. Best Practices

• •Mocking: Use unittest.mock to isolate external dependencies (e.g., file system, git operations).
• •Integration Tests: Place slower tests that touch the disk or inspect multiple modules (like caching tests) in separate files (e.g., test_integration_*.py).
• •No Side Effects: Tests must use temporary directories (tempfile) and clean up after themselves.