sap-cap

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

• Third-party content exposure detected (high risk: 0.85). The required workflow instructs the agent to call mcp__cap__search_docs/mcp__cap__search_model, which at runtime fetches CAP MCP server documentation/model text (i.e., third-party content not authored by the operating user) and injects it into the agent’s LLM context.