review-and-fix
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes structured findings from the CodeRabbit CLI, which could potentially contain malicious content intended to manipulate the agent's automated code fixes.\n
- Ingestion points: The skill parses the output from
coderabbit review --agentin Step 3.\n - Boundary markers: A multi-step validation process (Step 3.2) is implemented to ensure findings are true positives before action is taken.\n
- Capability inventory: The skill utilizes the Edit tool to modify project files and executes shell commands for testing and verification (Step 4, 5).\n
- Sanitization: The instruction set mandates logical verification and surrounding context checks to filter incorrect or malicious suggestions.- [COMMAND_EXECUTION]: The skill runs various shell commands to manage the code review lifecycle.\n
- Evidence: It executes
coderabbitCLI for version checks, authentication, updates, and reviews. It also performsgit fetchand runs project build, lint, and test scripts (e.g.,pnpm lint,tsc,pnpm test). These are standard operations for the tool's stated purpose.
Audit Metadata