reality-verification

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md command mapping explicitly instructs the agent to use MCP Playwright and the WebFetch tool to fetch/verify live sites and APIs (e.g., "site, page, live | MCP fetch/playwright" and "endpoint, API | WebFetch tool"), which causes the agent to ingest and act on arbitrary public third‑party content.