Fix Failing Tests Automatically

The problem

Debugging test failures is a time-consuming distraction that breaks development flow. Manually tracing stack traces to source code often leads to oversized fixes or unnecessary test suppressions.

What it does

• Analyzes test error output to map failures to specific source lines.
• Identifies root causes by cross-referencing test assertions with implementation logic.
• Applies the smallest possible safe fix to source code or configuration.
• Verifies resolutions by automatically rerunning the affected test suite.
• Generates a structured summary of modifications and verification results.

Why this beats prompting it yourself

Generic prompts often lead to agents deleting failing tests or performing broad refactors that introduce new bugs. This skill enforces a strict "source-first" fix logic and requires verification runs, ensuring the suite passes because the code works, not because the tests were weakened.

Use cases

• Fixing unit tests that broke after a dependency update or refactor.
• Debugging integration test failures in CI/CD pipelines.
• Resolving configuration issues in test runners or environment setups.
• Restoring passing status to a legacy test suite with minimal code changes.