test-mcp-server

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

• Third-party content exposure detected (medium risk: 0.30). In the required inspector fixture workflow, renderTool(name, input?, options?) calls the configured MCP server with attacker-controlled arguments (or uses server/simulation tool responses) and then renders/reads the resulting InspectorResult.content/structuredContent and optionally takes screenshots, so outsider-authored tool/resource text from any user-supplied server can be ingested at runtime.