Content
<div align="center">
<img src="assets/logo/Creative PromptX Duck Logo 4.svg" alt="PromptX Logo" width="120" height="120"/>
<p><strong>Professional Capability Enhancement System for Native AI Applications</strong></p>
<p>Providing professional roles, memory management, and knowledge systems for AI applications like Claude Desktop through the MCP protocol</p>
<p>
<strong><a href="README.md">中文</a></strong> |
<a href="README_EN.md">English</a> |
<a href="https://github.com/Deepractice/PromptX/issues">Issues</a>
</p>
</div>
<br/>
## 🚀 One-Click Start - Direct Connection for AI Applications
### 🔌 **MCP Standardized Access**
# Configure MCP Connection in AI Applications (Minimal Configuration)
#### **⚡ Zero Configuration Access (Recommended)**
Add the following to the MCP configuration file of the AI application:
```json
{
"mcpServers": {
"promptx": {
"command": "npx",
"args": ["-y", "-f", "dpml-prompt@snapshot", "mcp-server"]
}
}
}
```
**🎯 It's that simple!** PromptX will automatically:
- ✅ **Intelligently Identify Working Directory** - Automatically find the appropriate workspace
- ✅ **Automatically Initialize Environment** - No need to manually create folders and configurations
- ✅ **Dynamically Adapt to Projects** - Works seamlessly across different projects
#### **🔧 Advanced Configuration (Optional)**
For custom configurations, you can add the following options:
```json
{
"mcpServers": {
"promptx": {
"command": "npx",
"args": ["-y", "-f", "dpml-prompt@snapshot", "mcp-server"],
"env": {
"PROMPTX_WORKSPACE": "/your/custom/workspace/path"
}
}
}
}
```
### Not sure how to use MCP? [Click to view the MCP Kindergarten Tutorial on BiliBili](https://www.bilibili.com/video/BV1HFd6YhErb)
##### 🔧 Explanation of Advanced Options:
- `PROMPTX_WORKSPACE`: Custom workspace path (optional, system will auto-detect)
##### 📂 Format for Custom Workspace Path
- **Windows**: `"D:\\username\\path\\your\\project"` (use double backslashes)
- **macOS/Linux**: `"/Users/username/path/your/project"`
#### **AI Applications Supporting MCP**
| AI Application | Status | Configuration File Location | Features |
|----------------|--------|-----------------------------|----------|
| **Claude Desktop** | ✅ Official Support | Windows: `%APPDATA%\Claude\claude_desktop_config.json`<br/>macOS: `~/Library/Application Support/Claude/claude_desktop_config.json` | Official client from Anthropic, native MCP support |
| **Cursor** | ✅ Supported | Configured via MCP settings panel | Intelligent code editor, developer-friendly |
| **Claude Code** | ✅ Supported | `/home/user/.claude.json` or `~/.claude.json` | Official CLI tool from Anthropic, native MCP support, command-line AI programming assistant |
| **Windsurf** | ✅ Supported | MCP configuration panel within IDE | AI-native IDE launched by Codeium |
| **Cline** | ✅ Supported | VS Code plugin configuration | Powerful AI programming assistant |
| **Augment** | ✅ Supported | Desktop application configuration | AI-native code editor |
| **Trae** | ✅ Supported | IDE plugin configuration | AI-driven code generation and refactoring tool |
| **Tongyi Lingma** | 🟡 Planned Support | Alibaba Cloud IDE plugin | AI programming assistant launched by Alibaba Cloud |
| **Zed** | ✅ Supported | Configuration file: `~/.config/zed/settings.json` | High-performance code editor |
| **Continue** | ✅ Supported | VS Code plugin configuration | VS Code AI assistant plugin |
| **Replit Agent** | 🟡 Experimental Support | Built-in on Replit platform | Online programming environment |
| **Jan** | 🟡 In Development | Local AI client | Privacy-first local AI assistant |
| **Ollama WebUI** | 🟡 Community Support | Third-party MCP adapter | Local large model interface |
| **Open WebUI** | 🟡 Community Support | Plugin system | Open-source AI interface |
| **Baidu Comate** | 🟡 Planned Support | Baidu IDE plugin | AI programming assistant launched by Baidu |
| **Tencent CodeWhisperer** | 🟡 Planned Support | Tencent Cloud IDE | Tencent Cloud AI programming tool |
> **Note**:
> - ✅ **Official Support**: Native support for MCP protocol
> - 🟡 **Experimental/Community Support**: Supported through plugins or experimental features
> - More AI applications are being integrated with the MCP protocol...
**🎯 After configuration, AI applications will gain 6 professional tools:**
- `promptx_init` - 🏗️ System Initialization
- `promptx_hello` - 👋 Role Discovery
- `promptx_action` - ⚡ Role Activation
- `promptx_learn` - 📚 Knowledge Learning
- `promptx_recall` - 🔍 Memory Retrieval
- `promptx_remember` - 💾 Experience Saving
📖 **[Complete MCP Integration Guide](docs/mcp-integration-guide.md)**
### **📸 Usage Effects After Successful Configuration**
#### **🎭 Role Discovery and Activation**

*Step 1: Use `promptx_hello` to discover all available professional roles*

*Step 2: Use `promptx_action` to activate the professional role with one click, gaining full professional capabilities*
#### **💭 Intelligent Memory Management**

*Step 3: Use `promptx_remember` to save important information, and the AI will actively remember and recall it at the right moment*
> **✨ Configuration Tip**: After completing the MCP configuration, your AI application will gain all the above professional features. No additional learning is required; simply follow the prompts on the interface to enjoy professional AI services.
## ⭐ Star Growth Trend
[](https://star-history.com/#Deepractice/PromptX&Date)
### **Contribution Guidelines**
- 📋 **[Contribution Process](CONTRIBUTING.md)** - Detailed contribution guidelines and coding standards
- 🌿 **[Branching Strategy](docs/BRANCHING.md)** - Branch management and release process
- 🚀 **[Release Process](docs/RELEASE.md)** - Version management and release documentation
Scan the QR code to join the technical discussion group:
<img src="assets/qrcode.jpg" alt="Technical Discussion Group" width="200">
## 📄 License
MIT License - Making AI professional capabilities accessible
---
**🚀 Experience it now: Start the PromptX MCP Server and enhance your AI application with professional capabilities!**