use-dom
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by documenting how to render untrusted data in a WebView container.
- Ingestion points: The documentation shows props like 'content', 'code', and 'data' being passed directly into DOM-based components.
- Boundary markers: No examples demonstrate the use of delimiters or instructions to distinguish data from potentially malicious commands.
- Capability inventory: The skill describes methods for bridging native actions such as 'showAlert' and 'saveData' to the WebView.
- Sanitization: There is no mention or example of data sanitization to prevent script execution within the DOM context.
- [DYNAMIC_EXECUTION]: The skill uses the 'use dom' directive to execute web-based scripts in a native application container.
- Evidence: The 'use dom' directive is applied to components in SKILL.md to trigger isolated runtime execution in a WebView.
Audit Metadata