rust-learner

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly launches background agents that use agent-browser/WebFetch to fetch and parse content from public third‑party sites (crates.io, lib.rs, docs.rs, releases.rs, and explicitly "Reddit, TWIR, blogs"), so the agent will read and interpret untrusted/user‑generated web content as part of its workflow.