The Agent Skills Directory

[SAFE]: Framework detection is performed using grep and jq on local configuration files like package.json and requirements.txt, which is a standard and safe practice for development tools.
[SAFE]: The execution of test commands is protected by a whitelist in case blocks, ensuring that only expected commands like npm test or pytest are run, even if environment variables were manipulated.
[SAFE]: The skill implements local caching in the .claude/cache/ directory to improve performance, avoiding unnecessary file processing and network requests.
[SAFE]: TDD health monitoring uses git log to track local changes and find to calculate test-to-code ratios, remaining entirely within the local repository's boundaries.

tdd-red-green