requirements
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use
git config user.nameto automatically populate the author field in the requirements document. This is a benign use of a local configuration command for metadata purposes.- [INDIRECT_PROMPT_INJECTION]: The process involves reading user-provided research and existing documentation (REQUIREMENTS.md). While this represents an ingestion point for untrusted data, the skill is constrained to writing documentation and lacks sensitive capabilities such as network access or arbitrary code execution that would make this vector exploitable.- [SAFE]: All identified behaviors align with the stated purpose of generating a Product Requirements Document (PRD). The use of a static template inreferences/requirements.mdensures consistent and predictable output structure.
Audit Metadata