security-basics
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to analyze external, untrusted code diffs and files, creating an ingestion point for potentially malicious instructions. 1. Ingestion points: Untrusted source code and diffs provided by the user for review in SKILL.md. 2. Boundary markers: The skill does not implement explicit delimiters or safety instructions to isolate the untrusted code from the agent's primary instructions. 3. Capability inventory: The skill itself generates text-based reviews and code fixes, but the intended audience is a general coding agent which typically possesses file-system modification capabilities. 4. Sanitization: No specific input sanitization or filtering is defined in the skill logic.
Audit Metadata