What problem does it solve?
AI-generated code often passes the happy path while hiding technical debt: missing error handling, leaked resources, unhandled edge cases, and imports of packages or APIs that do not exist. This Skill forces a targeted audit of the specific failure patterns AI agents produce, catching debt before it reaches production.
Core Features & Use Cases
- Failure Mode Audit: Checks whether network timeouts, null inputs, and permission errors are handled with specific catches rather than swallowed exceptions.
- Orphan Detection: Verifies that every opened resource (connections, listeners, timers, temp files) has a corresponding cleanup call.
- Hallucinated Dependency Check: Confirms that every import exists in package.json or requirements.txt and that referenced API methods are real.
- Use Case: After an AI coding session generates 200 lines across several files, run this audit before merging the PR to surface swallowed errors, missing cleanup, and invented library methods.
Quick Start
Audit the code just generated in this session for hidden failure modes, orphaned resources, and hallucinated dependencies before I merge it.
