tech-spec
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data from the local project environment.
- Ingestion points: The process involves reading several local files, including
docs/ux-spec.md,AGENTS.md,CLAUDE.md, and general codebase files (components, APIs, routes). - Boundary markers: There are no explicit markers (like XML tags or unique delimiters) defined to separate instructions from the data being read from local files.
- Capability inventory: The skill allows the agent to create and write to files (specifically
docs/technical-spec.md). - Sanitization: The skill contains self-mitigating instructions, explicitly telling the agent to "Treat AI output as untrusted" and to validate results and protect sensitive data, though it does not implement technical sanitization of the input files themselves.
Audit Metadata