The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses common development CLI tools including npm, pnpm, npx, pyright, ruff, and git. These commands are used for standard project maintenance tasks such as building the project, running tests, and checking for linting or type errors.
[SAFE]: The security scan phase uses grep to identify potential hardcoded secrets (like OpenAI keys starting with 'sk-') within the local project files. This is a defensive coding practice intended to help the developer clean up credentials before committing code, and does not involve exfiltrating this data.

verification-loop