videodb

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly ingests and processes arbitrary public/untrusted content (e.g., coll.upload(url="https://..."), YouTube uploads, coll.connect_rtstream(url="rtmp://...") in SKILL.md and reference/rtstream*.md) and runs LLM-powered indexing/index_visuals/index_audio and alerting pipelines (capture & rtstream WebSocket events like "visual_index" and "audio_index") whose outputs are read and used to drive alerts, search, timeline edits, and follow-up actions—so untrusted third-party content can materially influence agent behavior.