data-viz
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides structured guidance for data visualization using established libraries like Recharts, D3, and shadcn/ui.
- [SAFE]: Implements a 'Dynamic color safety' principle to ensure text readability when using externally sourced color palettes, and provides explicit warnings against insecure template string interpolation when generating programmatic dashboards.
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and visualization of external datasets, creating a potential surface for indirect prompt injection if data contents are treated as instructions by the agent.
- Ingestion points: Raw data inputs for charts and dashboards (referenced in SKILL.md and references/component-guide.md).
- Boundary markers: None explicitly defined to isolate data values from system prompts.
- Capability inventory: Code generation and execution (React/JS, Python), file-system writes for dashboard files, and network operations for fetching datasets.
- Sanitization: Includes programmatic checks for color contrast and guidance on separating data from logic in HTML templates.
Audit Metadata