scrutinize
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to ingest and trace logic within untrusted external inputs such as pull requests, code changes, and design documents. This creates a potential attack surface where malicious instructions hidden in the code under review could influence the agent's behavior.
- Ingestion points: Untrusted source code files, PR diffs, and project plans mentioned throughout SKILL.md.
- Boundary markers: Absent. The skill does not define specific delimiters or "ignore embedded instructions" warnings for the external content being processed.
- Capability inventory: The workflow requires the agent to trace code paths end-to-end, which necessitates the use of file system reading and potentially codebase search tools.
- Sanitization: Absent. No instructions are provided for sanitizing or escaping content from external files before the agent processes it.
Audit Metadata