DartMCP

by G1Joshi

317 downloads
Not rated
GitHub

About

DartMCP is a Dart implementation of the Model Context Protocol (MCP) for AI agent interactions, providing both server and client components for building interoperable AI/LLM tools and services. It is designed for developers who want to integrate Dart applications with…

Details

Author
G1Joshi
Downloads
317
Categories
Other, AI

- Implements MCP protocol over standard I/O or custom channels
- Full server and client components in Dart 3.9.2+
- Supports logging, prompts, resources, and tools
- Cross-platform (macOS, Linux, Windows)
- Client streams for real-time server notifications

Add the dart_mcp package (version ^0.3.3) to your pubspec.yaml, run dart pub get, then implement a server using the Server class from lib/src/server.dart or a client using the Client class from lib/src/client.dart. Run the server with dart run path/to/your_server.dart or use the client example in example/dartmcp_example.dart.

DartMCP

A Dart implementation of the Model Context Protocol (MCP) for AI agent interactions.

Overview

DartMCP is a Model Context Protocol (MCP) implementation in Dart, providing both server and client components for building interoperable AI/LLM tools and services.

Requirements

- Dart SDK 3.9.2+
- Works on any platform supported by Dart (macOS, Linux, Windows)

Project Structure

The project is organized into two main components:

- lib/src/server.dart: The server implementation
- lib/src/client.dart: The client implementation

Dependencies

- MCP Dart SDK

Installation

1. Add DartMCP to your project:

dependencies:
  dart_mcp: ^0.3.3

2. Install dependencies:

dart pub get

Server

The DartMCP server component provides a Model Context Protocol server implementation in Dart. It enables AI agents and tools to interact with Dart applications using the standard MCP protocol.

Features

- Built with Dart 3.9.2+
- Implements MCP protocol over standard I/O or custom channels
- Includes support for logging, prompts, resources, and tools
- Cross-platform support

Getting Started

1. Implement your server logic using Server from lib/src/server.dart.
2. Run your server as a Dart application:

dart run path/to/your_server.dart

Server Component Details

The server component is responsible for:

- Handling incoming MCP connections
- Managing model context and state
- Processing client requests
- Maintaining state synchronization

Location: lib/src/server.dart

Client

The DartMCP client component allows Dart applications to connect to and interact with MCP-compatible AI agents and servers.

Features

- Built with Dart 3.9.2+
- Implements MCP client over standard I/O or custom channels
- Simple API for initializing and communicating with MCP servers
- Streams for server notifications (tools, resources, logs, etc.)

Getting Started

1. Use the Client class from lib/src/client.dart in your Dart application.
2. Example usage:

import 'package:dartmcp/dartmcp.dart';

void main() {
var client = Client();
client.main();
}

Client Component Details

The client component provides:

- Connection management to MCP servers
- Model context synchronization
- Real-time updates handling
- Client-side state management

Location: lib/src/client.dart

Example

See example/dartmcp_example.dart for a minimal usage example.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.