bridgebound-history-16

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

• Third-party content exposure detected (high risk: 0.70). The required workflow relies on ColdIQ MCP/API calls (e.g., mcp__coldiq__search_jobs, mcp__coldiq__find_signals, mcp__coldiq__find_people) to fetch job/signal text at runtime, which is outsider-authored free-form content coming from an external marketplace/vendor API and thus can be injected into the LLM context.