Desktop Extensions (DXT) provides a standardized format for packaging and distributing local MCP servers, enabling one-click installation within desktop applications. This specification, along with a CLI tool and core loading logic, aims to simplify the integration of AI tools and services into various desktop environments, fostering a more connected and extensible AI ecosystem.

How It Works

DXT packages are essentially zip archives containing an MCP server and a manifest.json file. This manifest describes the server's capabilities, ensuring compatibility and discoverability. The format is inspired by popular extension models like Chrome and VS Code, abstracting away the complexities of server setup and communication for end-users. This approach promotes portability and ease of adoption for local AI services across different applications.

Quick Start & Requirements

• Install CLI: npm install -g @anthropic-ai/dxt
• Create extension: Navigate to your MCP server directory and run dxt init, then dxt pack.
• Installation: Open the generated .dxt file with a compatible application like Claude for macOS/Windows.
• Full specification: MANIFEST.md
• Examples: examples

Highlighted Details

• Enables one-click installation of local MCP servers.
• Supports automatic updates and easy configuration of server variables.
• Provides a CLI for generating manifest.json and packing .dxt files.
• Designed for broad adoption across multiple AI desktop applications.

Maintenance & Community

This project is maintained by Anthropic. Further community engagement details are not explicitly provided in the README.

Licensing & Compatibility

The repository does not explicitly state a license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

The README does not specify a license, which may impact commercial adoption. Details on supported MCP server protocols beyond the implied @modelcontextprotocol/sdk are not provided.