code-review
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon code review feedback from external human or automated sources, creating a vulnerability surface where malicious instructions could be processed.
- Ingestion points: External feedback comments mentioned in SKILL.md and references/code-review-reception.md.
- Boundary markers: The skill includes a manual verification protocol (READ, UNDERSTAND, VERIFY, EVALUATE) which requires the agent to critically evaluate input before acting.
- Capability inventory: The agent has access to git commands and arbitrary shell commands for verification (tests, builds).
- Sanitization: No automated filtering is present; the skill relies on the agent following the technical skepticism guidelines provided.
- [COMMAND_EXECUTION]: The skill requires the execution of shell commands to perform its duties.
- Git operations: Commands like 'git rev-parse' and 'git log' are used to extract commit information for review requests.
- Verification gates: The protocol in references/verification-before-completion.md requires the agent to identify and run the appropriate verification commands (e.g., test suites or build tools) for the current project context.
Audit Metadata