tech-review-doc
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of technical design documents provided by the user to generate technical reviews. This ingestion represents a surface for indirect prompt injection.
- Ingestion points: Reads user-provided design documents at a specified path (SKILL.md, Step 0).
- Boundary markers: The output is constrained to a specific markdown template (template.md) and focused on business-language translation.
- Capability inventory: The skill has file read and write permissions but no network access or shell execution capabilities.
- Sanitization: The instruction set enforces a manual approval step (HARD-GATE) in Step 1, requiring the user to verify the core content before the agent proceeds to generate diagrams or write the final file.
- [EXTERNAL_DOWNLOADS]: The skill contains a manual remediation instruction for the user to install a missing dependency:
npx skills add FuDesign2008/open-skills -g --skill clarifying-question-discipline --yes. This targets the author's own repository and is a standard administrative command for the platform's ecosystem, not an automated execution pattern.
Audit Metadata