specs-api
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from prototype HTML pages located in
docs/product/demo/. This creates a surface for indirect prompt injection where malicious instructions embedded in the prototypes could influence the agent's output. - Ingestion points: HTML files within the
docs/product/demo/directory. - Boundary markers: The skill does not define specific delimiters or instructions to the subagent to disregard potential instructions found within the prototype data.
- Capability inventory: The agent and its subagents have read access to local project documentation and write access to create or merge specification files in the
docs/specs/directory. - Sanitization: The skill lacks descriptions for sanitizing or validating the input HTML content before it is parsed by the subagents.
Audit Metadata