bellingcat-osint

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

• Third-party content exposure detected (high risk: 0.70). The workflow instructs fetching a tool’s guide from a public GitBook URL at runtime (outsider-authored web content), which would be ingested as readable text into the agent’s LLM context.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent at runtime to fetch external guidance from https://bellingcat.gitbook.io/toolkit/more/all-tools/.md, which could supply remote instructions that directly influence agent behavior.