elevenlabs-agents

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill explicitly allows adding web URLs to the knowledge base ("Step 3: Add Knowledge Base (RAG) — PDFs, text files, web URLs") and documents that the agent "automatically searches knowledge base during conversation", and also includes webhook/server tools that fetch arbitrary external URLs (references/tool-examples.md and API reference), so untrusted third‑party content can be ingested and materially influence agent decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The widget-embed-template includes and will execute remote JavaScript at runtime from https://elevenlabs.io/convai-widget/index.js, which is a required external script dependency that runs code in the client and therefore can directly control the agent UI/behavior.