The Agent Skills Directory

[SAFE]: The skill implements a comprehensive review workflow using three distinct expert lenses to evaluate code quality, requirements fulfillment, and testing thoroughness.
[SAFE]: It properly integrates with vendor-specific tools get_context and get_impact to retrieve project context, ensuring the review is grounded in existing architectural patterns and documented principles.
[SAFE]: No malicious behavioral patterns, such as unauthorized data transmission, credential theft, or hidden command execution, were identified within the skill instructions.

perform-review