01Module 2: Model Context ProtocolA module map of the Model Context Protocol - from core concepts through architecture, primitives, building servers, security, ecosystem, and comparison with function calling.02What is MCP?The Model Context Protocol - announced by Anthropic in November 2024 - solves the N×M integration problem by giving AI systems a standard way to connect to any tool or data source.03MCP Architecture - Client-ServerDeep dive into MCP's client-server architecture - Host, Client, and Server roles; stdio and HTTP+SSE transport layers; JSON-RPC 2.0 message format; initialization handshake; capability negotiation; and full lifecycle.04MCP Tools, Resources, and PromptsDeep dive into MCP's three primitives - Tools (callable functions), Resources (readable data), and Prompts (reusable templates) - with complete Python implementations of each.05Building an MCP ServerHands-on guide to building a production-quality MCP filesystem server in Python using the official MCP SDK - complete with 4 tools, resources, MCP Inspector testing, and Claude Desktop integration.06MCP Security and PermissionsSecurity model of the Model Context Protocol - attack surfaces including tool poisoning, resource injection, and confused deputy attacks, plus permission scoping, transport security, and a production security checklist.07MCP Ecosystem and ServersThe growing MCP ecosystem - official Anthropic servers, community landscape, MCP registries, evaluating third-party servers, IDE integrations, and patterns for building ecosystem vs. team-specific servers.08MCP vs Function CallingDeep architectural comparison of MCP and function calling - where each operates, when to use each, the decision matrix, hybrid patterns, and how to migrate from function calling to MCP.
