The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides utility scripts (scripts/code-review-analyzer.py, scripts/pr-template-enforcer.sh, scripts/review-checklist.sh, scripts/review-comments.py) meant to be executed for automating analysis tasks. These scripts perform safe operations such as reading file content and generating report strings.
[PROMPT_INJECTION]: As the skill is designed to analyze user-provided code and PR descriptions, it inherently possesses a surface for indirect prompt injection.
Ingestion points: Source code files and PR description files processed by scripts/code-review-analyzer.py and scripts/pr-template-enforcer.sh.
Boundary markers: No specific boundary markers or instructions to ignore embedded commands are present in the scripts.
Capability inventory: The skill is restricted to reading local files for analysis; it does not have capabilities for network access, file system modification, or execution of arbitrary code from the analyzed files.
Sanitization: Content is processed as text or parsed via ast.parse (safe) without specific content sanitization, but the lack of dangerous capabilities mitigates this risk.
[CREDENTIALS_UNSAFE]: Multiple documentation files (examples/CODE-REVIEW-EXAMPLES.md, references/review-guide.md) contain example API keys such as sk_live_abc123xyz. These are clearly documented as placeholders and negative examples to demonstrate insecure coding practices that should be flagged during reviews.

code-review