The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's metadata and documentation include references to the author's own GitHub repository (github.com/jacob-balslev/skill-graph) for hosting configuration, protocol definitions, and evaluation artifacts.
[COMMAND_EXECUTION]: The skill is configured to use standard system tools including 'Read', 'Grep', and 'Bash'. These are used within the workflow to perform tasks such as finding call sites for modified functions or checking project conventions, which is expected behavior for a code review tool.
[SAFE]: The skill provides explicit instructions to detect security vulnerabilities (e.g., SQL injection, XSS, auth bypass) during the review process, treating them as high-severity blockers. It also mandates a 'no-rubber-stamp' policy for AI-generated code to prevent the introduction of security regressions.

code-review