taste
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides comprehensive guidelines for code quality, focusing on principles like KISS, SoC, and SLAP. No malicious code, obfuscation, or unauthorized network operations were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and audit external source code, creating a potential surface for indirect prompt injection.
- Ingestion points: Project source code files and symbols identified in the user-specified lane.
- Boundary markers: Not explicitly defined within the skill documentation.
- Capability inventory: Static analysis, reporting of violations, and performing behavior-preserving code edits as requested by the user.
- Sanitization: The skill relies on standard agent parsing of code files without additional sanitization layers.
Audit Metadata