The Agent Skills Directory

[SAFE]: The skill consists of instructional markdown files that guide an AI agent on how to perform thorough code reviews across five dimensions: correctness, readability, architecture, security, and performance.
[SAFE]: Includes detailed reference checklists for security (security-checklist.md) and performance (performance-checklist.md), which promote industry-standard best practices such as input validation, secret management, and efficient resource loading.
[SAFE]: No executable scripts or suspicious network operations are included within the skill's own logic. It provides commands for standard development tools (e.g., npm audit, lighthouse, webpack-bundle-analyzer) intended to be run by the developer or agent during the review process.
[SAFE]: The instructional content explicitly warns against security risks like hardcoded secrets, injection vulnerabilities, and insecure design, aligning with safe development practices.

code-review-and-quality