Model Context Protocol (MCP)

What is MCP?

The Model Context Protocol is an open standard (created by Anthropic) that provides a universal way for LLMs to connect with external data sources, tools, and services. Think of it as "USB for AI" — a standard interface that works across different models and applications.

Architecture

┌──────────┐     ┌──────────────┐     ┌──────────────┐
│  LLM App │────▶│  MCP Client  │────▶│  MCP Server  │
│ (Claude) │     │              │     │  (Your API)  │
└──────────┘     └──────────────┘     └──────────────┘
                                             │
                                      ┌──────┴──────┐
                                      │             │
                                   Tools       Resources

Key Concepts

• Tools: Functions the LLM can call (like API endpoints). Have input schemas and return results.
• Resources: Read-only data the LLM can access (like files, database records, documentation).
• Prompts: Pre-defined prompt templates with parameters, provided by the server.
• Transport: stdio (local) or HTTP with SSE (remote). Servers can run locally or as web services.

Why MCP Matters

• Standardization: Write once, use with any MCP-compatible client
• Ecosystem: Growing library of pre-built MCP servers for databases, APIs, file systems
• Security: Structured permissions and capability negotiation
• Composability: Connect multiple MCP servers to give an LLM access to many systems

Example: This Project Uses MCP!

The DaisyDo FastAPI service includes an MCP server that exposes Odoo's data through standard MCP tools. Any MCP-compatible client can search records, create entries, and manage the ERP system through natural language.

🌼 Daisy+ in Action: A Real MCP Implementation

Daisy+ is a real-world MCP implementation. The FastAPI service acts as an MCP server that bridges Claude (or any MCP-compatible AI) with the Odoo ERP. This means you can connect Claude Desktop, Cursor, or any MCP client directly to your business data — ask questions about sales, create records, generate reports — all through the standardized MCP protocol. Daisy+ proves that MCP isn't just a spec — it's a practical bridge between AI and enterprise software.