idea-refine
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a structured workflow for ideation and refinement. It includes multiple reference documents for software engineering best practices which are informative and benign.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input to generate content.
- Ingestion points: User-provided ideas via $ARGUMENTS and codebase context via Glob, Grep, and Read tools.
- Boundary markers: None present in the instructional text.
- Capability inventory: File reading, file writing to the docs/ideas/ directory, and interactive user questioning.
- Sanitization: None observed for the ingested content.
- [COMMAND_EXECUTION]: The skill includes a shell script (scripts/idea-refine.sh) that creates a local documentation directory. The script uses mkdir -p and contains no risky commands or network access.
Audit Metadata