harden
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions and code examples focus on software engineering best practices, specifically adding error handling, input validation, and logging to existing code to improve robustness.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided code, creating an indirect injection surface. Evidence chain: 1. Ingestion points: User code provided to the agent for hardening. 2. Boundary markers: None specified in the skill body. 3. Capability inventory: No tools are invoked by the skill itself, though the execution environment typically provides file and terminal access. 4. Sanitization: The skill explicitly directs the implementation of input validation and checks for injection ('Safe from injection?').
Audit Metadata