a2a-mcp-integration

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly describes agents that perform web search and document analysis (see "Scenario 2: Distributed Research Assistant" — "Search agent uses MCP web search tools" and "Analysis agent uses MCP document processing tools"), which indicates the agent will fetch and read untrusted public web/document content as part of its workflow and that those results drive subsequent agent actions.