mcp usecase

English | 한국어

Introduction

Quick-start Auto MCP is a tool that helps you easily and quickly register Anthropic's Model Context Protocol (MCP) in Claude Desktop and Cursor.

Key advantages: 1. Quick Setup: Add MCP functionality to Claude Desktop and Cursor simply by running a tool and copying/pasting the generated JSON file. 2. Various Tools Provided: We continuously update useful MCP tools. Stay up to date with your personalized toolkit by starring and following us. :)

Table of Contents

• Features
• Project Structure
• Requirements
• Installation
• Configuration
• Usage
• Troubleshooting
• License
• Contributing
• Contact
• Author

Features

• RAG (Retrieval Augmented Generation) - Keyword, semantic, and hybrid search functionality for PDF documents
• Dify External Knowledge API - Document search functionality via Dify's external knowledge API
• Dify Workflow - Execute and retrieve results from Dify Workflow
• Web Search - Real-time web search using Tavily API
• Automatic JSON Generation - Automatically generate MCP JSON files needed for Claude Desktop and Cursor

Project Structure

.
├── case1                     # RAG example
├── case2                     # Dify External Knowledge API example
├── case3                     # Dify Workflow example
├── case4                     # Web Search example
├── data                      # Example data files
├── docs                      # Documentation folder
│   ├── case1.md           # case1 description ? Includes tips for optimized tool invocation
│   ├── case2.md           # case2 description
│   ├── case3.md           # case3 description
│   ├── case4.md           # case4 description
│   └── installation.md    # Installation guide
├── .env.example              # .env example format
├── pyproject.toml            # Project settings
├── requirements.txt          # Required packages list
└── uv.lock                   # uv.lock

Requirements

• Python >= 3.11
• Claude Desktop or Cursor (MCP supporting version)
• uv (recommended) or pip

Installation

1. Clone the repository

git clone https://github.com/teddynote-lab/mcp.git
cd mcp

2. Set up virtual environment

Using uv (recommended)

# macOS/Linux
uv venv
uv pip install -r requirements.txt

# Windows
uv venv
uv pip install -r requirements_windows.txt

Using pip

python -m venv .venv

# Windows
.venvScriptsactivate
pip install -r requirements_windows.txt

# macOS/Linux
source .venv/bin/activate

pip install -r requirements.txt

3. Preparing the PDF File

Plese prepare a PDF file required for RAG in the ./data directory.

Configuration

In order to execute each case, a .env file is required. Please specify the necessary environment variables in the .env.example file located in the root directory, and rename it to .env.

sites for configuring required environment variables for each case

• https://platform.openai.com/api-keys
• https://dify.ai/
• https://app.tavily.com/home

Usage

1. Generate JSON File

Run the following command in each case directory to generate the necessary JSON file:

# Activate virtual environment

# Windows
.venvScriptsactivate

# macOS/Linux
source .venv/bin/activate

# Navigate to example directory
cd case1

# Generate JSON file
python auto_mcp_json.py

2. Register MCP in Claude Desktop/Cursor

1. Launch Claude Desktop or Cursor
2. Open MCP settings menu
3. Copy and paste the generated JSON content
4. Save and restart

Note: When you run Claude Desktop or Cursor, the MCP server will automatically run with it. When you close the software, the MCP server will also terminate.

Troubleshooting

Common issues and solutions:

• MCP Server Connection Failure: Check if the service is running properly and if there are no port conflicts. In particular, when applying case2, you must also run dify_ek_server.py.
• API Key Errors: Verify that environment variables are set correctly.
• Virtual Environment Issues: Ensure Python version is 3.11 or higher.

License

MIT LICENSE

Contributing

Contributions are always welcome! Please participate in the project through issue registration or pull requests. :)

Contact

If you have questions or need help, please register an issue or contact: [email protected]

Author

Hantaek Lim