decision-sheet
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests text from the conversation history to populate questions and options in a generated HTML artifact. This introduces a surface where malicious content from the chat could be rendered to the user.
- Ingestion points: User-influenced data is extracted by the agent and written to
questions.jsonin the scratchpad. - Boundary markers: The skill does not explicitly define delimiters for the ingested data.
- Capability inventory: The skill is capable of writing files to the workspace and publishing artifacts to the platform's UI.
- Sanitization: The skill provides robust sanitization; the
build.mjsscript encodes script-sensitive characters like<and the HTML template uses a dedicatedescfunction to handle HTML entity escaping for all dynamic content. - [EXTERNAL_DOWNLOADS]: The HTML template references Google Fonts. These are well-known and trusted external resources used for styling the user interface.
Audit Metadata