codebase-analysis
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): The skill is purely instructional and consists of markdown documentation and pattern templates. No executable scripts, obfuscated code, or persistence mechanisms are present.- [INDIRECT_PROMPT_INJECTION] (LOW): The skill is designed to process untrusted external data in the form of codebases. Ingestion points: Local files accessed via Read, Grep, Glob, and LSP tools. Boundary markers: None provided in the instructions. Capability inventory: Read, Grep, Glob, and LSP (all limited to read-only operations). Sanitization: None performed on the content of analyzed files. The severity remains low because the skill lacks the ability to execute code, write to the filesystem, or communicate over the network.
Audit Metadata