iterating

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly reads user-pasted/uploads and project knowledge and notes that "RAG system may also inject relevant content into context" (references/chat-environment.md), so the agent will parse and act on potentially untrusted third-party/user-generated content mounted at /mnt/project or injected by RAG.