d3-viz
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides visualization templates that ingest external data and render it directly to the document using methods that do not provide automatic sanitization. \n
- Ingestion points: Data objects passed to the
drawVisualizationfunction inSKILL.md, and thedataprops in theBasicChart(assets/chart-template.jsx) andInteractiveChart(assets/interactive-template.jsx) components. \n - Boundary markers: None identified in the code snippets or instructions to warn the agent about untrusted data content. \n
- Capability inventory: The skill uses
d3.selection.html()to render data-derived strings in tooltips (found inSKILL.mdandassets/interactive-template.jsx). This method allows for the injection of arbitrary HTML and JavaScript into the browser context if the data labels or values contain malicious payloads. \n - Sanitization: No sanitization, validation, or escaping of the data labels or values is performed before they are interpolated into the HTML strings.\n- [EXTERNAL_DOWNLOADS]: The skill includes instructions to load the D3.js library from
https://d3js.org/d3.v7.min.jsusing a CDN link. This targets a well-known service for the library's distribution.
Audit Metadata