Ai Validation

This is a local code analysis tool that runs validation checks against AI-generated code to catch the kinds of problems LLMs tend to miss. It scans your codebase for test bias (where tests only validate happy paths), identifies edge cases you haven't covered, and flags gaps in test coverage that could hide bugs. Reach for this when you're using Claude or another AI to write code and want an independent second pass that checks whether the tests are actually robust or just optimistic. The stdio transport means it runs locally, so your code never leaves your machine. Think of it as a sanity check layer between AI output and production.