The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the agent to execute standard repository verification tools, such as ruff for linting and pytest for running tests, as part of the pull request review process. This is a core part of its intended functionality for ensuring code quality.
[SAFE]: The provided engineering rules specifically address and promote security best practices, including secure handling of JWTs, the use of httpOnly cookies for authentication, and the prevention of secret exposure in frontend code.
[SAFE]: References to external packages and model repositories (e.g., Silero VAD on GitHub, or standard Python/Node libraries) target well-known and reputable services within the technology community.
[SAFE]: The skill includes comprehensive architectural guidelines designed to minimize code complexity and improve maintainability through principles like deep modules and the deletion test.

collective-pr-review