review-prd
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted PRD documents to identify technical claims for verification against the codebase. A malicious document could contain crafted strings to influence search operations or skip verification steps.
- Ingestion points: PRD file content provided via the
prdinput. - Boundary markers: Absent; the agent extracts raw strings from the document.
- Capability inventory: Grep, Glob, Git, and File Write (Edit) tools.
- Sanitization: Absent; codebase searches are performed using content extracted directly from the PRD.
- [COMMAND_EXECUTION]: The skill uses local Node.js and shell scripts for document resolution and ID management. These scripts are provided within the skill package and do not perform remote code execution.
Audit Metadata