The Agent Skills Directory

[SAFE]: The skill defines structured criteria for code reviews across three roles (Critic, Security, and Architect). These guidelines assist the AI in identifying security and quality issues in plans and source code.
[SAFE]: The provided scripts (check-complexity.sh, pre-review-check.sh, run-linters.sh) are standard utility tools that perform local static analysis. They utilize common Unix utilities like find, grep, and awk, as well as official language linters (ESLint, Prettier, Ruff, Go vet) to detect code smells and formatting issues.
[INDIRECT_PROMPT_INJECTION]: The skill is intended to process external source code and planning documents, which represents a potential attack surface for indirect prompt injection. However, the associated analysis scripts only perform read-only static checks and linting, which are inherently low-risk activities. Standard agent guardrails for processing external data are expected to be in place.
[EXTERNAL_DOWNLOADS]: The run-linters.sh script utilizes npx to execute tools like ESLint and Prettier. This is standard developer behavior that fetches packages from the official NPM registry when local versions are unavailable, following established development practices.

code-review-checklist