
MCP practices in AI coding | MCP 在 AI 编程中的实践
概览
is a community-maintained collection of Model Context Protocol (MCP) servers designed for the editor. By integrating MCP servers, TRAE can break through the limitations of pure text interaction, connecting to local file systems, databases, browsers, and various third-party services, becoming a powerful all-around development assistant. : This project primarily hosts . If you are looking for for TRAE, please check out TRAE Agent Skills. is an open standard protocol introduced by Anthropic, designed to standardize how AI models connect with external systems. If the AI model is the "brain", then MCP is its "senses" and "limbs". It empowers AI with the following capabilities: - : Execute command lines, send messages, manage code repositories, etc. - : Access local files, query databases, read documents, etc. - : Seamlessly interact with external platforms like Slack, GitHub, Google Drive, etc. In TRAE, configured MCP servers serve as callable for the agent.
README
以下为原文 README 节选,完整内容请查看原文。
TRAE MCPs
TRAE MCPs is a community-maintained collection of Model Context Protocol (MCP) servers designed for the TRAE editor.
By integrating MCP servers, TRAE can break through the limitations of pure text interaction, connecting to local file systems, databases, browsers, and various third-party services, becoming a powerful all-around development assistant.
💡 Note: This project primarily hosts MCP Servers (Tools). If you are looking for Skills (SOPs) for TRAE, please check out TRAE Agent Skills.
📖 What is MCP?
Model Context Protocol (MCP) is an open standard protocol introduced by Anthropic, designed to standardize how AI models connect with external systems.
If the AI model is the “brain”, then MCP is its “senses” and “limbs”. It empowers AI with the following capabilities:
- 🔧 Operate Tools: Execute command lines, send messages, manage code repositories, etc.
- 📂 Read Data: Access local files, query databases, read documents, etc.
- 🌐 Connect Services: Seamlessly interact with external platforms like Slack, GitHub, Google Drive, etc.
In TRAE, configured MCP servers serve as callable Tools for the agent. The agent can automatically select and execute these tools based on task requirements.
🚀 Features
- Out of the Box: Provides MCP servers that are tested and adapted for the TRAE environment.
- Community Driven: Gathers wisdom from community developers to continuously expand the tool library.
- Unified Standard: Follows the MCP standard protocol to ensure compatibility and stability.
📦 MCP Server List
| MCP Name | Description | Status | Documentation |
|---|---|---|---|
| CloudBase | Tencent CloudBase MCP — AI models, auth, databases, cloud functions, storage, CloudRun (npx @cloudbase/cloudbase-mcp) |
✅ Ready | README |
| (To be added) | More MCP servers are under development… | 🚧 | - |
(The repository is currently initializing. PRs to contribute your recommended MCPs are welcome!)
🛠️ How to Use
To use the MCP servers provided in this repository in TRAE, please follow these steps:
-
Open TRAE MCP Settings
- Open TRAE.
- Navigate to
Settings>MCP.
-
Configure MCP Config Click Add - Manually Add, and add the server configuration:
{ "mcpServers": { "your-mcp-name": { "command": "node", "args": ["/absolute/path/to/trae-mcp/mcp/your-mcp-name/build/index.js"], "env": { "API_KEY": "your-api-key" } } } }
安装
This server does not publish a one-line install command.
Open the repository installation guide 新手?先看 mcpskill.org 教程 →配置
{
"mcpServers": {
"your-mcp-name": {
"command": "node",
"args": ["/absolute/path/to/trae-mcp/mcp/your-mcp-name/build/index.js"],
"env": {
"API_KEY": "your-api-key"
}
}
}
}