deep-research

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs the agent to fetch and crawl open web content using the webfetch tool and to search public GitHub repositories with gh_grep, meaning it ingests untrusted, user-generated third‑party content (web pages, forums, and repo content) and is expected to read and interpret that content as part of its workflow.