microsoft-agent-framework

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

• Third-party content exposure detected (high risk: 0.90). This skill's docs and workflows explicitly describe accepting and incorporating untrusted third-party/tool results—e.g., Model Context Protocol (references/mcp.md) and AG-UI frontend/backend tool flows (references/official-docs/integrations/ag-ui/* and getting-started.md) where remote MCP servers and client-provided tool results are fetched and fed back into the agent run—so external, user-controlled content can influence agent decisions.