generative-thinking

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md "Random stimulus" section explicitly tells an LLM agent to source external content (e.g., "fetch a random Wikipedia article" or "pull a noun from a URL the user has open"), which requires ingesting public/user-edited web content that could carry instructions and thus influence subsequent agent decisions.