infographic-generator
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (JSON, CSV, or text descriptions) to generate graphics, which presents a surface for potential indirect instructions to influence agent behavior.
- Ingestion points: External data sources are ingested during the 'Probe' and 'Execute' phases (SKILL.md).
- Boundary markers: The workflow includes 'Constrain' and 'Evidence' phases to validate data, but lacks explicit markers to prevent the execution of embedded instructions within the data.
- Capability inventory: The skill utilizes Python execution (Matplotlib), Mermaid.js rendering, and file system write operations to save generated HTML and image files.
- Sanitization: The 'Verify' step focuses on representation accuracy and readability rather than sanitizing malicious patterns or escaping the content.
- [DYNAMIC_EXECUTION]: The skill involves the dynamic generation and execution of code and markup (HTML, Python/Matplotlib, Mermaid.js) based on runtime user data.
- Evidence: The 'Execute' section describes generating self-contained HTML/CSS and using Python to produce PNG/SVG charts, which are subsequently saved to the local directory ('Deliver' phase).
Audit Metadata