🌟 CloudBase AI ToolKit

Such as Cursor | WindSurf | CodeBuddy etc. Click to see list of supported AI development tools

Visit Tencent CloudBase Console to create an environment. New users can start for free.

Make sure Node.js v18 or higher is installed on your computer. You can download and install the latest version from Node.js official website.

To improve dependency download speed, it's recommended to set npm registry to Tencent mirror. You can run the following command in terminal:

npm config set registry https://mirrors.cloud.tencent.com/npm/

This speeds up dependency downloads, especially in mainland China.

Run the following command in terminal:

npx -y clear-npx-cache 

Step 1: Use AI Rules

Template includes .cursor/rules/ directory, AI will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. Please click the popup window at the bottom left to enable MCP. If not starting from template, click the button below to install in Cursor:

Or manually add configuration to .cursor/mcp.json:

{
  "mcpServers": {
    "cloudbase-mcp": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Then click the ⚙️ icon in the top-right corner of Cursor client, go to "MCP" settings, find cloudbase and enable it.

Step 3: Switch to Agent Mode

Use Agent mode in the chat window for code generation and automation.

Step 1: Use AI Rules

Template's .windsurf/ directory contains configurations optimized for WindSurf. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in WindSurf's Plugins configuration:

{
  "mcpServers": {
    "cloudbase-mcp": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Step 3: Switch to Write Mode

Switch to Write mode in chat for intelligent generation.

Step 1: Use AI Rules

Template includes .rules/ directory, CodeBuddy will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

1. You can click the MCP button in the top-right corner of CodeBuddy, search for CloudBase in the MCP marketplace, and click install.

2. Or manually add: click the MCP button in the top-right corner of CodeBuddy, click add on the right, and add CloudBase MCP in the MCP configuration:

{
  "mcpServers": {
    "cloudbase": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Step 3: Switch to Craft Agent

Switch to Craft mode in the chat window. It's recommended to disable the confirm plan feature in the right settings for a smoother experience.

Step 1: Use AI Rules

Template includes .clinerules/ directory, AI will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in MCP Server configuration:

{
  "mcpServers": {
    "cloudbase": {
      "autoApprove": [],
      "timeout": 60,
      "command": "npx",
      "args": [
        "@cloudbase/cloudbase-mcp@latest"
      ],
      "transportType": "stdio",
      "disabled": false
    }
  }
}

Step 3: Use AI Chat

It's recommended to use models that support code generation and function calls.

Step 1: Use AI Rules

Template's .github/ directory contains Copilot optimized configurations. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in .vscode/mcp.json:

{
    "servers": {
        "cloudbase": {
            "command": "npx",
            "args": [
                "@cloudbase/cloudbase-mcp@latest"
            ]
        }
    }
}

Step 3: Switch to Agent Mode

Switch to Agent mode in the bottom-left corner of the chat window.

This guide supports both Trae international and Trae CN versions. It's recommended to test with Claude/DeepSeek V3 0324 models.

Step 1: Use AI Rules

Template's .trae/rules directory contains CloudBase rule configurations for Trae. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

Add in Trae's MCP configuration:

{
  "mcpServers": {
    "cloudbase-mcp": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Step 3: Use Builder with MCP Chat

Select Builder with MCP in the agent for chat.

Step 1: Use AI Rules

Template includes .lingma/ directory, Tongyi Lingma will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in MCP configuration:

{
  "mcpServers": {
    "cloudbase-mcp": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Step 3: Switch to Agent Mode

Switch to agent mode in the bottom-left corner of the chat window.

Step 1: Use AI Rules

Template includes .roo/rules directory, RooCode will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in MCP configuration:

{
    "mcpServers": {
        "cloudbase": {
            "command": "npx",
            "args": [
                "@cloudbase/cloudbase-mcp@latest"
            ],
            "disabled": false
        }
    }
}

Step 3: Use AI Chat

Use the chat window for code generation and automation.

Step 1: Use AI Rules

Template includes .comate/rules directory, Baidu Comate will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, add in MCP configuration:

{
    "mcpServers": {
        "cloudbase": {
            "command": "npx",
            "args": [
                "@cloudbase/cloudbase-mcp@latest"
            ],
            "disabled": false
        }
    }
}

Step 3: Use Agent Mode

Use Agent mode in the chat window for intelligent code generation and automation.

Step 1: Use AI Rules

Template includes .augment-guidelines file, Augment Code will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

If using template project, MCP configuration is pre-configured (built into .vscode/settings.json). If not starting from template, add CloudBase MCP in Augment's MCP configuration, refer to documentation:

{
  "augment.advanced": {
     "mcpServers": {
        "cloudbase": {
            "command": "npx",
            "args": [
                "@cloudbase/cloudbase-mcp@latest"
            ]
        }
    }
  }
}

Step 3: Use Agent Mode

Use Agent mode in the chat window for intelligent code generation and automation.

Step 1: Use AI Rules

Template includes CLAUDE.md file, Claude Code will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules.

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, create .mcp.json file in project root:

{
  "mcpServers": {
    "cloudbase": {
      "command": "npx",
      "args": [
        "-y",
        "@cloudbase/cloudbase-mcp@latest"
      ]
    }
  }
}

Step 3: Use AI Chat

Chat directly with AI in Claude Code for intelligent code generation and automation.

Step 1: Install Gemini CLI

Make sure Node.js 18+ is installed, then install Gemini CLI globally:

npm install -g @google/gemini-cli

Or run directly:

npx @google/gemini-cli

On first run, log in with your personal Google account to get free quota (60 requests per minute, 1000 requests per day).

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured.

If not starting from template, create .gemini/settings.json file in user home directory (~) or project directory:

{
  "mcpServers": {
    "cloudbase-mcp": {
      "command": "npx",
      "args": ["-y", "@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

Step 3: Enable AI Rules

Template project root already includes GEMINI.md file with CloudBase development rules.

If it's an existing project, tell the AI in Gemini CLI:

Download CloudBase AI rules in current project

Step 4: Start Development

Launch Gemini CLI:

gemini

Then tell the AI:

Login to CloudBase

Common Commands:

• /tools - View available tools
• /mcp - View MCP server status
• /memory show - View current rules
• !command - Execute Shell command

Step 1: Install OpenAI Codex CLI

Make sure Node.js 18+ is installed, then install OpenAI Codex CLI globally:

npm install -g @openai/codex

Or run directly:

npx @openai/codex

Step 2: Configure MCP

[!TIP] If the tool count remains 0 after installation, please refer to FAQ

If using template project, MCP configuration is pre-configured. If not starting from template, there are two configuration methods:

Method 1: Use project configuration file (recommended)

Create .codex/config.toml file in project root:

[mcp_servers.cloudbase]
command = "npx"
args = ["-y", "@cloudbase/cloudbase-mcp@latest"]

Then start with the specified configuration file:

codex --config .codex/config.toml

Method 2: Global configuration

Create ~/.codex/config.toml file in user home directory:

[mcp_servers.cloudbase]
command = "npx"
args = ["-y", "@cloudbase/cloudbase-mcp@latest"]

Step 3: Enable AI Rules

Template includes AGENTS.md file, OpenAI Codex CLI will automatically recognize CloudBase best practices. If not starting from template, you can ask AI to help download CloudBase rules:

Download CloudBase AI rules in current project

Step 4: Start Development

Launch OpenAI Codex CLI:

codex

If using project configuration file:

codex --config .codex/config.toml

Chat with AI in Codex CLI:

Login to CloudBase

Then you can start developing your requirements.

Update AI Rules

If you want to update to the latest CloudBase AI rules in an existing project, simply tell the AI:

Download CloudBase AI rules

AI will automatically download and update the latest rule configurations to your project directory.

Update MCP Tools

When new versions of MCP tools are released, you can update through:

1. Auto Update (Recommended): In your AI development tool's MCP list, find cloudbase-mcp and re-enable or refresh the MCP list to automatically install the latest version

2. Manual Update: If auto update doesn't work, you can disable and re-enable cloudbase-mcp, or restart your AI IDE

Since the MCP configuration uses the @latest tag, it usually automatically gets the latest version.