superpowers-receiving-code-review
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines behavioral and procedural guidelines for code reviews and does not contain malicious code or instructions.- [PROMPT_INJECTION]: The instructions are designed to refine the agent's professional persona and task-specific logic; they do not contain patterns aimed at bypassing core AI safety guardrails.- [COMMAND_EXECUTION]: The skill references the use of
grepand the GitHub CLI (gh api) for code searching and PR interactions, which are appropriate for the stated task.- [INDIRECT_PROMPT_INJECTION]: The skill specifically mitigates indirect injection risks by instructing the agent to treat external feedback as untrusted data that must be verified against the codebase and architectural constraints before implementation.
Audit Metadata