mcp-dev-tools

An MCP (Model Context Protocol) server that integrates AI into software development workflows. Connects directly to Claude Desktop and any MCP-compatible client.

What It Does

This server exposes 3 AI-powered development tools via the Model Context Protocol:

Once connected to Claude Desktop, you can invoke these tools in natural language:

• "Give me a Python code review checklist"

• "Generate a commit message for: added JWT auth to the login endpoint"

• "Format the docs for my validate_patient_record function"

Related MCP server: Forja MCP Server

Architecture

┌─────────────────────┐     JSON-RPC / stdio     ┌──────────────────────────┐
│   Claude Desktop    │ ◄─────────────────────── │   mcp-dev-tools server   │
│   (MCP Client)      │ ──────────────────────── │   (Python / FastMCP)     │
└─────────────────────┘                          │                          │
                                                 │  * code_review_checklist │
         OR                                      │  * generate_commit_msg   │
                                                 │  * api_docs_formatter    │
┌─────────────────────┐                          └──────────────────────────┘
│  Your Python Client │ ◄── stdio_client() ──────────────────┘
│  (client.py)        │
└─────────────────────┘

Getting Started

1. Install dependencies

git clone https://github.com/SKeval/mcp-dev-tools
cd mcp-dev-tools
pip install -r requirements.txt

2. Test with the included client

python client.py

Expected output:

mcp-dev-tools connected. Available tools:
   * code_review_checklist
   * generate_commit_message
   * api_docs_formatter
...
All 3 tools working. mcp-dev-tools is ready.

3. Connect to Claude Desktop

Add this to your Claude Desktop config file:

Mac: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "mcp-dev-tools": {
      "command": "python",
      "args": ["/absolute/path/to/mcp-dev-tools/server.py"]
    }
  }
}

Restart Claude Desktop. You'll see mcp-dev-tools appear in the tools panel.

Tool Reference

code_review_checklist

Returns a structured checklist for code reviews.

Example:

result = await session.call_tool(
    "code_review_checklist",
    {"language": "python", "include_general": True}
)

generate_commit_message

Generates a Conventional Commits message.

Example output:

feat(fhir): added FHIR resource validation to the patient data ingestion pipeline

Type: feat — A new feature
Scope: fhir

api_docs_formatter

Converts a raw docstring into structured Markdown API docs.

Why MCP?

Traditional AI integrations require custom one-off code for every tool connection. MCP standardizes this — build the server once, connect it to any MCP-compatible client. This is the same pattern applicable to clinical software development: a single MCP server can expose AI-assisted code review, compliance checking, and documentation generation across an entire development team's workflow.

License

MIT License