Introduction to Programming with Python
Google AI Professional Certificate - Learn AI Skills That Get You Hired
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn how to represent agents as MCP servers using the mcp-agent framework in this 18-minute conference talk. Discover how this approach enables agents to be run from any MCP client, including Claude and Cursor, through the Model Context Protocol. Explore the paradigm shift from traditional "agentic" behavior existing only on the client side to agents functioning as MCP servers themselves, allowing any MCP client to invoke, coordinate, and orchestrate agents like any other MCP server. Understand the four key benefits this approach provides: agent composition for building complex multi-agent systems over the same base protocol, platform independence for using agents across any MCP-compatible client, scalability for running agent workflows on dedicated infrastructure, and customization for developing reusable agent workflows. Examine how mcp-agent combines Anthropic's Model Context Protocol standardized interface with their Building Effective Agents patterns into a composable AI application framework that manages MCP server connection lifecycles and implements production-ready agent patterns in a chainable way.
Syllabus
Exposing Agents as MCP servers with mcp-agent: Sarmad Qadri
Taught by
AI Engineer