cap-de-jsapi-filters
Pass
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill provides instructions for capturing and acting upon data sent from external Domo iframes via the MessagePort API, which represents an indirect prompt injection vulnerability surface.\n
- Ingestion points: Data enters the application via the
window.addEventListener('message', ...)listener and theMessagePort.onmessagehandler inSKILL.md.\n - Boundary markers: Absent. The instructions do not include directions to treat incoming iframe data as untrusted or to use delimiters.\n
- Capability inventory: The skill suggests using the ingested data for significant actions, including client-side routing (
router.push) and triggering further RPC calls to iframes.\n - Sanitization: The provided code examples lack origin verification (
event.origincheck) and do not sanitize or validate the structured data (filters, appData) received from the remote iframe source.
Audit Metadata