Test and Fix

You can automate pytest unit tests and fix failing code by running the skill from your project root, which auto-detects changes, executes targeted test suites, proposes root-cause fixes, and re-runs tests until stability is achieved.

The best way to diagnose unit test failures and apply code fixes is to use an automated workflow that runs targeted tests, analyzes root causes, applies minimal code fixes, and re-runs tests to verify stability.

Yes, you need a project-level configuration file at .claude/skill-config.md to run automated tests and debugging, which allows the tool to target changes and execute tests properly.

Yes, you can run specific targeted tests instead of the full suite, as the skill supports targeted test selection based on changes detected in your repository to streamline debugging.

The automated test runner handles code fixes by proposing minimal fixes for diagnosed failures and automatically re-running the affected tests until stability is achieved, providing clear failure reporting.

If your unit tests keep failing after automated code fixes, review the clear failure reporting provided by the test runner to manually diagnose root causes that may fall outside the automated iteration loop.