huashu-keynote-black
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a controlled environment for generating presentation artifacts. It references well-known external resources (Google Fonts) and developer repositories neutrally, consistent with its design-oriented purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a generation engine that interpolates user-provided text (topics, business metrics, and strategy priorities) into an HTML template. Ingestion points: User input provided via prompt and the 'topic', 'audience', and 'accent' fields in open-design.json. Boundary markers: The instructions mandate a 'replace content only' workflow within the example.html template. Capability inventory: The skill uses file writing (fs:write) to output the generated HTML artifact. Sanitization: No specific escaping or sanitization instructions are defined for the user-provided content before it is placed into the document DOM, representing a standard surface for document-generating agents.
Audit Metadata