MCP Prompt Server

🚀 MCP Prompt Server

The MCP Prompt Server offers pre - set prompt templates, enabling tasks like code review, API documentation generation, and code refactoring. It provides flexible templates with dynamic parameter replacement and allows developers to customize templates. Additionally, it offers a tool API and is optimized for editors such as Cursor and Windsurf.

✨ Features

• Provide preset prompt templates for tasks like code review, API documentation generation, and code refactoring.
• Offer all prompt templates as MCP tools instead of the MCP prompts format.
• Support dynamic parameter replacement to make prompt templates more flexible.
• Allow developers to freely add and modify prompt templates.
• Provide a tool API to reload prompts and query available prompts.
• Optimized for editors like Cursor and Windsurf to provide a better integration experience.

📦 Installation

Install Dependencies

cd prompt-server
npm install

Start the Server

npm start

The server will run on standard input/output and can be connected to by Cursor, Windsurf, or other MCP clients.

💻 Usage Examples

Add New Prompt Templates

You can create new prompt templates by adding new YAML or JSON files in the src/prompts directory. Each template file should contain the following content:

name: prompt_name                # Unique identifier for invoking this prompt
description: prompt description  # Description of the prompt's function
arguments:                       # List of arguments (optional)
  - name: arg_name               # Argument name
    description: arg description # Argument description
    required: true/false         # Whether it is required
messages:                        # List of prompt messages
  - role: user/assistant         # Message role
    content:
      type: text                 # Content type
      text: |                    # Text content, can contain parameter placeholders
        {placeholder}

Tool API

The server provides the following tool APIs:

• getPrompts(): Get all available prompt templates.
• executePrompt(promptName, parameters): Execute the specified prompt template and pass parameters.
• addPrompt(promptTemplate): Add a new prompt template.
• removePrompt(promptName): Delete the specified prompt template.

Integration with Editors

Cursor Integration

1. Install the MCP plugin in Cursor.
2. Configure the plugin and add the following server information:

{
  "mcpServers": {
    "prompt-server": {
      "command": "node",
      "args": [
        "/path/to/prompt-server/src/index.js"
      ],
      "transport": "stdio"
    }
  }
}

3. After adding, restart Cursor and refresh the tool panel.

Windsurf Integration

1. Add the following content to the ~/.codeium/windsurf/mcp_config.json file:

{
  "mcpServers": {
    "prompt-server": {
      "command": "node",
      "args": [
        "/path/to/prompt-server/src/index.js"
      ],
      "transport": "stdio"
    }
  }
}

2. After adding, restart Windsurf and refresh the tool panel.

📚 Documentation

Directory Structure

prompt-server/
├── package.json         # Project dependencies and scripts
├── src/                 # Source code directory
│   ├── index.js         # Server entry file
│   └── prompts/         # Directory for preset prompt templates
│       ├── code_review.yaml
│       ├── api_documentation.yaml
│       ├── code_refactoring.yaml
│       ├── test_case_generator.yaml
│       └── project_architecture.yaml
└── README.md            # Project documentation

Expansion Suggestions

• Add more prompt templates for specialized fields: Such as natural language processing, machine learning, etc.
• Implement prompt version control: Support different versions of prompt templates.
• Add prompt classification and tags: Facilitate management and search.
• Implement prompt usage statistics and analysis: Record usage frequency and effectiveness.
• Add a user feedback mechanism: Collect user evaluations and suggestions for prompts.