Llm Gateway MCP Server

🚀 [Project Introduction]

This is an open - source project based on the Model Context Protocol (MCP), aiming to provide a unified interaction interface for various AI models. The project integrates multiple functional modules, including file processing, network requests, data parsing, database operations, etc.

🚀 Quick Start

Clone the repository

git clone https://github.com/your-repository.git
cd your-repository

Install dependencies

pip install -r requirements.txt

Start the service

python main.py

Configuration options

You can configure the following parameters in the config.json file:

{
  "host": "localhost",
  "port": 8000,
  "model_provider": "openai",
  "api_key": "your-api-key"
}

Usage Examples

Basic Usage

Call the service via the command line:

curl http://localhost:8000/api/ping

Advanced Usage

Use it in Python code:

import requests

response = requests.get('http://localhost:8000/api/ping')
print(response.json())

✨ Features

Core Features

• Multi - model support: Compatible with mainstream LLM providers such as OpenAI, Anthropic, Google, DeepSeek, xAI, etc.
• Tool extensibility: It has built - in various practical tools covering file operations, network requests, data processing, etc.
• Context management: Achieve effective interaction between models and external tools through a structured protocol.

Toolset

File System Tools

• file_read: Read the content of a file from the specified path.
• file_write: Write content to a file, supporting overwrite or append mode.
• dir_list: List all files and sub - directories in the specified directory.
• file_delete: Delete the specified file or empty directory.

Network Request Tools

• http_get: Send a GET request to get web page content.
• http_post: Send a POST request to submit data.
• http_request: A general interface supporting custom HTTP methods (e.g., PUT, DELETE, etc.).

Data Processing Tools

• json_parse: Parse JSON - formatted data.
• csv_read: Read a CSV file and convert it into structured data.
• xml_process: Process XML - formatted data, supporting XPath queries.

Database Operation Tools

• sql_query: Execute an SQL query and return the result set.
• db_connect: Establish a connection with the database.
• schema_parse: Parse and validate the database table structure.

Security Precautions

To ensure the security of the system, please follow the following principles during use:

1. Principle of least privilege: Run service processes with the lowest possible privileges to limit their access to sensitive resources.
2. Input validation: Strictly validate user - input data to prevent malicious code injection.
3. Log management: Avoid recording sensitive information such as API keys or user credentials in logs.

📚 Documentation

Overall Security Recommendations

• Regularly update project dependency libraries to prevent known security vulnerabilities.
• Configure appropriate access control policies to limit unnecessary network exposure.
• Use the HTTPS protocol for communication to protect the security of data during transmission.

Precautions for Tool Usage

• When handling file system operations, strictly limit the accessible directory range to prevent directory traversal attacks (e.g.,../).
• When executing SQL queries, give priority to using ORM frameworks or parameterized queries to avoid SQL injection risks.
• For operations that require executing arbitrary scripts (e.g., browser_evaluate_script), it is recommended to perform strict input validation and permission control.

📄 License

This project is licensed under the MIT license. For specific terms, please refer to the LICENSE file in the project.

Acknowledgments

This project relies on the support of many excellent open - source projects and technology communities. Special thanks to the following organizations and individuals:

• The Model Context Protocol (MCP) project team, who provided the basic conceptual framework and protocol specifications.
• The FastAPI team, who developed a high - performance Web framework.
• The Pydantic developers, who provided powerful tools for data validation and configuration management.
• The author of the Rich library, who brought a beautiful information terminal output experience.
• The uv project team, who provided a fast Python package installation solution.
• The Playwright team, who developed a powerful browser automation framework.
• The maintainers of OpenPyXL, who provided support for Excel file processing.
• The community contributors, thank you for your enthusiastic participation and continuous feedback.