MCP Server Toolkit

A Model Context Protocol (MCP) server that equips AI coding agents (Claude Code, Codex, Cursor, OpenCode, Gemini CLI) with filesystem, Git, database, and computation tools.

What is MCP?

The Model Context Protocol is an open standard that enables AI agents to securely interact with external tools and data sources. An MCP server speaks JSON-RPC 2.0 over stdio, exposing tools, resources, and prompts to any MCP-compatible client.

Related MCP server: MIST - Model Intelligence System for Tasks

Tools

Filesystem

Git

Database (SQLite)

Computation

Quick Start

# Install
pip install -r requirements.txt

# Configure your AI agent (Claude Code example)
# Add to .mcp.json:

{
  "mcpServers": {
    "toolkit": {
      "command": "python",
      "args": ["-m", "mcp_server.main"],
      "cwd": "/path/to/mcp-server-toolkit"
    }
  }
}

Environment Variables

Architecture

AI Agent (Claude Code / Codex / Cursor)
    |
    | JSON-RPC 2.0 over stdio
    v
MCPServer
    ├── Filesystem Tools  (read, write, list)
    ├── Git Tools         (status, log, diff, branch, show)
    ├── Database Tools    (query, list tables, schema)
    └── Calculator        (safe eval)

Tech Stack

• Python 3.10+ — No external dependencies

• JSON-RPC 2.0 — MCP protocol transport

• sqlite3 — Built-in database support

• Path security — Tools are sandboxed to MCP_SAFE_ROOTS

Part of the Hogan Dong Agent Stack

This project is the tool belt in a multi-agent platform:

AgentForge can plug this toolkit directly via MCP, giving its agents filesystem, Git, and database access.

Also: Browser MCP · RAG Agent · Agent Playground