textbook-grounding

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly performs parallel semantic searches across external/public sources (Φ2: "Multi-query semantic search across pdf-embeddings + web" and the pdf-search / pdf-brain / perplexity commands) and ingests those third‑party materials for thesis extraction and synthesis, exposing the agent to untrusted user-provided web/PDF content that could carry indirect prompt-injection.