brainstorming
Warn
Audited by Snyk on Jun 20, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.65). The runtime path
scripts/helper.jscaptures click text from the rendered HTML (target.textContent) and sends it over WebSocket as JSON (type: 'click'/type: 'choice'), which the server persists toSTATE_DIR/events; if the agent later reads these events and includes them in the LLM context, that click text is outsider-authored free text (any user interacting with the UI).
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata