build-brief
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a standard surface for indirect prompt injection by design, as it ingests user-provided project details and writes them into documentation files that may be read by subsequent AI agents.
- Ingestion points: User responses to interview questions as defined in the Workflow section of SKILL.md.
- Boundary markers: The templates provided in references/output-documents.md do not include explicit delimiters or instructions to treat user-derived content as untrusted data.
- Capability inventory: The skill performs file system writes (creation of .md and .png files) and utilizes an external image generation tool (imagegen).
- Sanitization: No explicit input sanitization or validation routines are specified before writing user content to the persistent project files.
Audit Metadata