Course Wiki

MCP

The Model Context Protocol: a standard way to connect agents to external tools and data sources.

MCP is the protocol layer for agent extensions: instead of hand-wiring each tool into each agent, a tool exposes itself as an MCP server (documentation lookup, database queries, issue trackers — the course uses a Context7-style documentation server as the example), and any MCP-capable agent discovers and calls it through the standard interface.

Module 5 teaches the capability, not one product: servers define tools, resources, and prompts; clients (coding agents) enumerate them and invoke them mid-task. The protocol also carries the security discussion — an MCP server is code the agent will execute, so trust, scoping, and audit matter — which is the module’s pairing of capability with constraint.

Taught in

• Machine Learning Zoomcamp

  • Module 1: Introduction to Machine Learning
    • Model Selection Process

• LLM Zoomcamp

  • Module 3: Orchestration
    • AI Agents

• AI Dev Tools Zoomcamp

  • Module 4: DevOps and Observability for AI-Built Apps
    • DevOps and Observability for AI-Built Apps
  • Module 5: Coding Agent Building Blocks: Reusable Skills and Specialized Subagents
    • Module 5 — Coding Agent Capabilities: MCP, Skills, Plugins, and Custom Agents
    • Configuration

• Data Engineering Zoomcamp

  • Module 5: Data Platforms
    • Getting Started with Bruin
    • Using Bruin MCP with AI Agents ## Related concepts

• Agent Skills and Subagents

• Coding Agents

• Context Engineering

• Git Worktrees