enhance-docs
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external documentation files which serve as untrusted data ingestion points.
- Ingestion points: Reads content from files in
docs/,agent-docs/, and project context files likeREADME.md,CLAUDE.md, andAGENTS.md(as defined in the 'Documentation Locations' section). - Boundary markers: The instructions do not define clear delimiters or use 'ignore embedded instructions' warnings when the agent processes these documentation files.
- Capability inventory: The skill defines a workflow involving 'Discover', 'Parse', and 'Fix' (file discovery, reading, and writing). The 'Auto-Fixes' section explicitly describes automated modifications to document structure and content.
- Sanitization: There is no logic provided to sanitize or filter out potential natural language instructions embedded within the target documentation files, which could lead to accidental obedience by the agent.
Audit Metadata