mom-test

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

• Third-party content exposure detected (high risk: 0.80). The skill's "Analyze Interview Responses" step explicitly instructs users to paste customer interview notes ("Here are notes from my customer interview: [paste notes]") for the agent to analyze, so the agent ingests arbitrary user-generated/untrusted third‑party content that could contain injected instructions.