The Agentic Protocol Stack

intermediate ~5 hours architecture mcp-servers agents transport

Understand the three open protocols forming the agentic AI communication stack:

  • MCP (agent-to-tool) — how agents use external tools and data
  • A2A (agent-to-agent) — how agents collaborate across platforms
  • AG-UI (agent-to-user) — how agents communicate with frontends

All three are converging under the Linux Foundation’s Agentic AI Foundation (AAIF). This path teaches you the architecture and practical usage of each layer.

Steps

  1. What is the Model Context Protocol (MCP)? mcp beginner

    MCP is the agent-to-tool layer — it standardizes how AI models discover and invoke external capabilities. Start here because MCP is the most mature of the three protocols and provides the foundation the other layers build on.

  2. Architecture overview mcp beginner

    The host/client/server three-tier architecture is MCP's core design. Understanding this separation is essential before comparing with A2A and AG-UI, which operate at different layers of the stack.

  3. Tools mcp intermediate

    Tools are the primary MCP primitive and the most direct connection to agent capabilities. Understanding how tools are defined via JSON Schema and invoked via the protocol helps you see what A2A and AG-UI complement.

  4. Understanding A2A a2a intermediate

    In-depth protocol explanation

    A2A addresses agent-to-agent communication where MCP addresses agent-to-tool access. The two protocols are complementary, not competing — an agent might use MCP to access tools and A2A to delegate subtasks to other agents on different platforms.

  5. Core architecture ag-ui advanced

    Understand how AG-UI connects front-end applications to AI agents

    AG-UI completes the stack by standardizing how agents communicate with user interfaces. While MCP connects to tools and A2A connects to agents, AG-UI connects to frontends — making it the protocol users actually see and interact with.

  6. Events ag-ui intermediate

    Understanding events in the Agent User Interaction Protocol

    AG-UI's typed event system (text deltas, tool calls, state updates) is what enables rich agent UIs beyond simple chat. Understanding the event types helps you design frontends that can visualize agent reasoning and tool usage in real time.

  7. MCP, A2A, and AG-UI ag-ui intermediate

    Understanding how AG-UI complements and works with MCP and A2A

    This is the synthesis page that shows how MCP, A2A, and AG-UI form a complete communication stack for agentic AI. All three are now under the Linux Foundation's AAIF, signaling industry convergence toward these standards.

  8. Build an MCP server mcp intermediate

    Get started building your own server to use in Claude for Desktop and other clients.

    Building an MCP server is the most hands-on way to understand the tool layer of the protocol stack. This quickstart gets you to a working server in minutes — focus on how tool registration maps to what you learned in the spec.

  9. Introduction ag-ui beginner

    Learn how to get started building an AG-UI integration

    This quickstart shows how to connect an agent backend to a React frontend using AG-UI events. It is the practical counterpart to the architecture concepts — focus on how events flow from agent to UI and how the frontend renders them.