ask-refactoring-readability
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill handles user-provided code, which is an untrusted ingestion point. 1. Ingestion points: User code provided via natural language triggers. 2. Boundary markers: Not specified in instructions; relies on implicit markdown structure. 3. Capability inventory: Restricted to reasoning and code suggestion; no automated execution or write access is granted. 4. Sanitization: None.
Audit Metadata