od-prototype-doc-kami-parchment
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates user-supplied text from the
$ARGUMENTSvariable and instructions returned by theprepare_open_design_brieftool into its operational workflow. - Ingestion points: User-provided arguments in
SKILL.mdand output from theopen-designMCP server tools. - Boundary markers: None identified; the skill directly interpolates user input into the brief and instructs the agent to follow instructions from the tool output without delimiters or safety warnings.
- Capability inventory: The skill explicitly directs the agent to use file-editing tools and register artifacts based on external instructions.
- Sanitization: There is no evidence of sanitization or validation of the brief content or the tool's returned instructions before they are used to author files.
Audit Metadata