explore-idea
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external repository content (initiative documents, existing code, and guidance) to generate an intent document, creating a surface for indirect instructions to influence the agent's output.
- Ingestion points: repository guidance, initiative documents, and existing code (processed in SKILL.md).
- Boundary markers: The skill partially mitigates this by instructing the agent to 'keep authoritative source references rather than copying whole documents' and to 'label proposed approaches and unverified assumptions distinctly from agreement.'
- Capability inventory: The skill performs file read operations on repository content and file write operations to 'docs/work//intent.md'. It explicitly prohibits code execution, commits, or dependency installation.
- Sanitization: There is no mention of technical sanitization or escaping for the ingested content before it is interpolated into the generated markdown template.
Audit Metadata