Content
# MCP Prompt Server (Go Version)
This is a MCP prompt server implemented in Go, based on the [mark3labs/mcp-go](https://github.com/mark3labs/mcp-go) library.
Inspired by [joeseesun/mcp-prompt-server](https://github.com/joeseesun/mcp-prompt-server) and [gdli6177/mcp-prompt-server](https://github.com/gdli6177/mcp-prompt-server), with initial prompts copied from [joeseesun/mcp-prompt-server](https://github.com/joeseesun/mcp-prompt-server).
## Have You Encountered These Pain Points?
Ever had a bunch of prompts but couldn't remember which one to use when?
Tired of copying and pasting prompts every time you need them?
Some people store prompts in AI programming tool rules, which solves part of the problem.
But what if we could turn common prompts into MCP tools?
By designing prompt templates as tools, we can call various prompts through natural language dialogue.
## The Power of This MCP
No more copying and pasting lengthy prompts.
Just use natural language dialogue to automatically:
- Generate visual web pages
- Design PRDs
- Create attractive titles
- And more...
AI will automatically find and use the appropriate prompts.
Suitable for any MCP-supported tool, such as Raycast, Cursor, Windsurf, Cherrystudio, etc.
## Main Features
- 📦 **Rich Prompt Templates**: Built-in high-quality code, writing, product, knowledge card, web generation, structured summary, and other prompts
- 🛠️ **Plug-and-Play MCP Tools**: All prompts automatically registered as MCP tools, supporting parameterized calls, compatible with mainstream editors
- 🔄 **Hot Reload and Management**: Reload new prompts without restarting the server
- 🧩 **Easy to Extend**: Add new YAML/JSON files to extend functionality without modifying core code
- 🏷️ **Multi-Language and Multi-Domain**: Suitable for Chinese/English content, product, education, media, AI, and other fields
## Directory Structure
```
mcp-prompt-server-go/
├── cmd/
│ └── main.go # Program entry
├── internal/
│ ├── server/ # Server implementation
│ │ └── server.go
│ ├── prompt/ # Prompt processing
│ │ ├── loader.go # Prompt loader
│ │ └── model.go # Prompt data structure
│ └── util/ # Utility functions
│ └── files.go
├── prompts/ # All prompt templates (YAML/JSON files)
│ ├── gen_summarize.yaml
│ ├── gen_title.yaml
│ └── ...
└── README.md
```
## Quick Start
1. **Clone the Repository**
```bash
git clone https://github.com/yourusername/mcp-prompt-server-go.git
cd mcp-prompt-server-go
```
2. **Build the Project**
```bash
./build.sh
```
3. **Start the Server**
```bash
./bin/mcp-prompt-server
```
The MCP prompt server will automatically load all prompt templates in the `prompts/` directory and expose them as MCP tools.
## Tool Integration
### Raycast
1. In Raycast, search for `install server (MCP)`
2. Give your MCP a simple name, such as `prompt` (for easy @ invocation later)
3. Command: Enter the path to the executable file after building, e.g., `/Users/yourusername/mcp-prompt-server-go/bin/mcp-prompt-server`
4. After saving, Raycast will automatically integrate the MCP prompt server
### Cursor
- Edit `~/.cursor/mcp_config.json` and add the following content (replace the path with your actual project path):
```json
{
"servers": [
{
"name": "Prompt Server",
"command": "/path/to/mcp-prompt-server-go/bin/mcp-prompt-server",
"args": [],
"transport": "stdio"
}
]
}
```
### Windsurf
- Edit `~/.codeium/windsurf/mcp_config.json` and add:
```json
{
"mcpServers": {
"prompt-server": {
"command": "/path/to/mcp-prompt-server-go/bin/mcp-prompt-server",
"args": [],
"transport": "stdio"
}
}
}
```
## How to Extend Prompts
1. **Create New YAML or JSON Files in the `prompts/` Directory**
2. **Template Example**:
```yaml
name: your_prompt_name
description: What this prompt does
arguments: []
messages:
- role: user
content:
type: text
text: |
Your prompt content, supports parameter placeholders like {{param}}
```
3. **Hot Reload Prompts**
- Use the `reload_prompts` tool in your editor or restart the server
## Management and Debugging
- `reload_prompts`: Hot reload all prompt templates
- `get_prompt_names`: List all available prompt names
## Frequently Asked Questions
- **Prompts Not Working?**
Check the YAML format, ensure the name field is unique, and reload or restart the service.
- **Parameters Not Working?**
Ensure the `arguments` field is correct and parameters are passed correctly.
## Contribution and Feedback
- Welcome to contribute new prompts, suggestions, and error reports!
## License
MIT
Connection Info
You Might Also Like
markitdown
Python tool for converting files and office documents to Markdown.
OpenAI Whisper
OpenAI Whisper MCP Server - 基于本地 Whisper CLI 的离线语音识别与翻译,无需 API Key,支持...
claude-flow
Claude-Flow v2.7.0 is an enterprise AI orchestration platform.
oh-my-opencode
Background agents · Curated agents like oracle, librarians, frontend...
ai-engineering-from-scratch
Learn it. Build it. Ship it for others. The most comprehensive open-source...
hyperframes
Write HTML. Render video. Built for agents.