icons
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
sheet.htmlfile fetches UI fonts from Google's official font service (fonts.googleapis.comandfonts.gstatic.com). These are well-known, trusted resources used for the skill's dashboard interface. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests untrusted SVG data from project files or user input (paste/drag-and-drop) and interpolates this data into prompts for the AI agent.
- Ingestion points: SVG markup is read from the local file system or provided by the user via the
sheet.htmldashboard's paste/drop functionality. - Boundary markers: The skill uses markdown triple-backtick fences to delimit SVG code within the generated prompts, which helps the agent distinguish data from instructions.
- Capability inventory: The agent is instructed to write or modify SVG files in the project directory based on the linting results.
- Sanitization: SVG markup is parsed through the browser's DOM parser (
template.innerHTML), which provides basic structural validation, though the raw markup is eventually passed back to the agent in the "Fix The Findings" prompts.
Audit Metadata