Ln 635 Test Isolation Auditor

The ln-635-test-isolation-auditor skill enables AI agents to systematically review codebases for test isolation issues, ensuring that individual tests do not depend on the state left by other tests. It identifies shared resources, global variable mutations, and non-deterministic dependencies that could cause flaky or unreliable test suites.

When to use it

• Before merging large batches of pull requests to prevent introducing hidden test coupling.
• During code refactoring when modifying shared utilities or database schemas.
• When debugging intermittent CI/CD pipeline failures caused by race conditions.
• To establish a baseline for test quality in new microservice architectures.

Key capabilities

• Detects shared state usage across independent test functions.
• Identifies global variable mutations affecting test order independence.
• Flags external resource dependencies like network calls or file system locks.
• Analyzes code structure to pinpoint non-deterministic logic within test suites.

Example prompts

• "Audit this repository for test isolation violations and list the specific functions causing conflicts."
• "Identify any global state mutations in the src/services directory that could break parallel test execution."
• "Review the recent changes in branch feature/auth-update and highlight tests that rely on shared database fixtures."

Tips & gotchas

Ensure your codebase includes sufficient metadata or comments describing test setup logic, as the auditor relies on analyzing actual code patterns rather than documentation. This skill is most effective when paired with a robust CI environment capable of running tests in parallel to verify the auditor's findings.