code-reviewer
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious code, hidden instructions, or unauthorized data access patterns were identified in the skill instructions or metadata.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to process untrusted code snippets.
- Ingestion points: External code provided by the user via review trigger phrases.
- Boundary markers: None defined; instructions do not provide delimiters for the input code.
- Capability inventory: No tools, scripts, or network capabilities are active or requested.
- Sanitization: None specified; the skill relies on the core agent's safety protocols.
Audit Metadata