{ "name": "yamcp", "version": "0.2.2", "description": "Organize and manage your MCP servers in YAM (Yet Another MCP) workspaces - create dedicated workspaces for Research, Coding, Data Analysis, and more", "types": "dist/index.d.ts", "repository": { "type": "git", "url": "https://github.com/hamidra/yamcp" }, "bin": { "yamcp": "dist/index.js" }, "scripts": { "dev": "tsx src/index.ts", "build": "pnpm run clean && tsc && pnpm run copy-example && chmod +x dist/index.js", "copy-example": "cp src/example-servers.json dist/example-servers.json", "inspect": "npx @modelcontextprotocol/inspector node dist/index.js", "clean": "rm -rf dist", "lint": "eslint src --ext .ts,.tsx", "test": "vitest run", "test:watch": "vitest", "start": "node dist/index.js", "prepublishOnly": "pnpm run build" }, "keywords": [], "author": "", "license": "MIT", "devDependencies": { "@modelcontextprotocol/inspector": "^0.10.2", "@types/node": "^20.17.30", "@types/prompts": "^2.4.9", "@types/treeify": "^1.0.3", "@types/uuid": "^9.0.8", "eslint": "^9", "typescript": "^5", "vitest": "^1.0.0" }, "dependencies": { "@modelcontextprotocol/sdk": "^1.9.0", "boxen": "^8.0.1", "chalk": "^5.4.1", "commander": "^13.1.0", "dotenv": "^16.4.7", "enquirer": "^2.4.1", "env-paths": "^3.0.0", "net": "^1.0.2", "open": "^10.1.0", "ora": "^8.2.0", "prompts": "^2.4.2", "treeify": "^1.1.0", "tsx": "^4.19.3", "uuid": "^9.0.1", "winston": "^3.17.0", "zod": "^3.24.2" }, "files": [ "dist" ] }

🍠 YAMCP - A Model Context Workspace Manager

YAMCP (YAM-C-P) is a command-line tool for organizing and managing MCP servers as local workspaces. It seamlessly connects to multiple MCP servers, local or remote, grouping them into a unified workspace exposed as Yet Another MCP server (YAM) for AI applications. You can create dedicated workspaces based on specific functionality (e.g., a YAM workspace for coding, design, research, ...) or based on the AI apps that consume servers (e.g., a YAM for Cursor, Claude, Windsurf) or any other combination in between. In addition, it simplifies monitoring and debugging MCP servers by centralizing all server communication logs in a single store, eliminating the need to dig through each AI client app’s logs separately.

Import and Create Workspaces

Connect All Bundled Servers in a Workspace to Your AI Apps with One Config

🚀 Quick Start

# Install YAMCP
npm install -g yamcp  # or use npx yamcp

# Import servers (choose one)
yamcp server import [config]    # import servers from config file (see src/example-servers.json for format)
yamcp server add                # or add manually

# create workspaces (e.g. a yam for coding, design, data, ...)
yamcp yam create

# Run workspace in your AI app
yamcp run <yam-workspace-name>

🔑 Key Concepts

• MCP Servers: Remote or local servers that provide Model Context Protocol services
• Workspaces (YAMs): Collections of MCP servers grouped together to be shared with AI Apps (e.g. a workspace for coding, writing, design, magic making!)
• Gateway: A local MCP server that manages connections to configured MCP servers in a workspace and exposes them through a unified server to AI App's MCP clients

With YAMCP, you can:

• Create workspaces to group MCP servers by AI application (e.g. Cursor, Claude, GitHub Copilot)
• Group servers by workflow purpose (e.g. software development, data science, technical writing)
• Connect AI apps to a single gateway that provides access to all workspace servers
• Manage and monitor multiple MCP server connections through a unified interface
• Track all server communications with detailed logging and debugging capabilities

Top-Level Commands

yamcp [command] [subcommand] [flags]

Available top-level commands:

• server - Manage MCP providers
• yam - Manage workspaces (yams)
• run - Run the gateway with a workspace
• log - View the server log location

🔧 Mcp Server Management Commands

Server Commands

yamcp server add      # Add a new MCP server (interactive)
yamcp server list     # List all configured servers and their status
yamcp server remove   # Remove a server configuration
yamcp server import   # Import server configurations from a JSON file

🍠 Yam Workspace Management Commands

Workspace Commands

yamcp yam create      # Create a new workspace (interactive)
yamcp yam list        # List all workspaces or show specific workspace details
yamcp yam edit        # Modify an existing workspace configuration
yamcp yam scan            # Scan workspaces
yamcp yam delete          # Delete a workspace

Runtime Commands

yamcp run <yam-workspace-name>   # Start the gateway with specified workspace
yamcp log                    # View server communication logs

✅ Command Reference

🖥️ YAMCP UI (created by @eladcandroid)

YAMCP UI provides an intuitive web interface to manage your MCP servers and workspaces through a universal dashboard.

# Run directly with npx (recommended)
npx yamcp-ui

# Or install globally
npm install -g yamcp-ui
yamcp-ui

Refer to the project repo for full documentation: https://github.com/eladcandroid/yamcp-ui.

🏗️ System Architecture

graph TB
    CLI[CLI Commands]
    GW[McpGateway]
    GS[GatewayServer]
    GR[GatewayRouter]
    LOG[Logger]
    STORE[(Store)]
    AI_APP[AI App]

    %% CLI Command Flow
    CLI -->|manages| STORE
    CLI -->|runs| GW

    %% Gateway Components
    GW -->|uses| GS
    GW -->|uses| GR
    GW -->|logs| LOG

    %% Server & Router
    GS -->|stdio transport| AI_APP
    GR -->|connects to| SERVERX
    GR -->|connects to| SERVERY

    %% Data Store
    STORE -->|loads config| GW

    %% External MCP Servers
    subgraph "Workspace Servers"
        SERVERX["Server x (Stdio)"]
        SERVERY["Server y (SSE)"]
    end
    %% Store Components
    subgraph "Configuration Store"
        PROVIDERS[(Provider Config)]
        WORKSPACES[(Workspace Config)]
    end
    STORE --- PROVIDERS
    STORE --- WORKSPACES

    classDef primary fill:#2374ab,stroke:#2374ab,color:#fff
    classDef secondary fill:#ff7e67,stroke:#ff7e67,color:#fff
    classDef store fill:#95b8d1,stroke:#95b8d1,color:#fff

    class GW,GS,GR primary
    class CLI,AI_APP secondary
    class STORE,PROVIDERS,WORKSPACES store

The diagram shows the main components of the YAMCP system:

• CLI Commands: User interface for managing servers and workspaces
• McpGateway: Core component that coordinates the Gateway Server and Router
• GatewayServer: Handles communication with AI Apps via stdio transport
• GatewayRouter: Manages connections to configured MCP servers
• Logger: Provides consolidated logging for all components
• Store: Manages configuration for providers and workspaces
• MCP Servers: Both local (stdio) and remote (SSE) servers that provide MCP services

🪪 Security Audits