visualize
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches versioned frontend libraries including Preact, HTM, and Mermaid.js from the https://esm.sh/ and https://cdn.jsdelivr.net/ CDNs to provide functionality for the generated visualizations.
- [DYNAMIC_EXECUTION]: Generates standalone HTML files containing JavaScript to render interactive content. The skill instructions mandate the use of Preact's tagged templates for automatic escaping and explicitly prohibit the use of innerHTML to prevent the execution of malicious logic from user-provided data.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied subjects and system descriptions to generate its output.
- Ingestion points: User-provided subjects, plans, and code flows in the input prompt.
- Boundary markers: Not explicitly defined in the raw input, though managed by the skill's templating logic.
- Capability inventory: Capability to write generated HTML files to the local file system and open them in a browser.
- Sanitization: Instructions strictly require the use of Preact's default escaping and forbid innerHTML to ensure untrusted content is not rendered as executable markup.
Audit Metadata