write-idea
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a product management tool that follows a strict, multi-step process for research and documentation. It implements security-positive ground rules, such as the exclusion of technical implementation details (separation of concerns) and a requirement for explicit user approval before any file is written to the disk.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests raw product ideas from the user and pending inbox entries from the repository. This intake of untrusted data constitutes an indirect prompt injection surface. However, the risk is mitigated by the skill's design, which uses interactive question-and-answer phases and requires the user to review and approve the final draft before it is saved to the specification registry.
- Ingestion points: User-provided idea descriptions and inbox entries in
inbox/<repository>/. - Boundary markers: Interactive resolution of the idea (Step 3) and a final draft review phase (Step 8).
- Capability inventory: File system write access (
docs/specs/) and market research via web search tools. - Sanitization: Relies on human-in-the-loop review and approval.
- [EXTERNAL_DOWNLOADS]: The skill instructions mandate market and competitive research using web search tools. This is a primary function for product discovery and does not involve downloading executable scripts or remote binaries.
Audit Metadata