Get the FREE Ultimate OpenClaw Setup Guide →

CloudBase

CloudBase MCP - Connect CloudBase to your AI Agent. Go from AI prompt to live app.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio tencentcloudbase-cloudbase-mcp npx @cloudbase/cloudbase-mcp@latest

How to use

CloudBase MCP acts as a deployment bridge between AI IDEs and Tencent Cloud Base, enabling you to transform prompts into deployed cloud resources without manual DevOps steps. The server leverages the CloudBase MCP toolkit to generate, deploy, and manage cloud resources such as cloud functions, databases, hosting, and other CloudBase services directly from your AI environment. You can run the MCP locally with npx for full functionality (including local file system interactions) or opt for hosted mode to run the MCP on Tencent Cloud, connecting via HTTP with your credentials. Use the provided plugin-enabled workflow to deploy web apps, games, or back-end services with automated configuration and one-click deployment. The server supports a guided flow from code generation to deployment, including automatic configuration of databases, security rules, and hosting setups, delivering access links once deployment completes.

How to install

Prerequisites:

  • Node.js installed on your machine (for local mode via npx).
  • Internet access to fetch MCP packages from npm.

Install and run locally (recommended for full feature set):

  1. Ensure Node.js and npm are installed on your machine.
  2. Run the MCP server via npx using the official package:
# Install and run the MCP locally
npx @cloudbase/cloudbase-mcp@latest
  1. In your AI IDE, configure the MCP server using the following payload:
{
  "mcpServers": {
    "cloudbase": {
      "command": "npx",
      "args": ["@cloudbase/cloudbase-mcp@latest"]
    }
  }
}

If you prefer hosted/云端 mode (no local Node.js required):

  1. Replace the local configuration with hosted mode details and credentials, for example:
{
  "mcpServers": {
    "cloudbase": {
      "type": "http",
      "url": "https://tcb-api.cloud.tencent.com/mcp/v1?env_id=<your_env_id>",
      "headers": {
        "X-TencentCloud-SecretId": "<your-secret-id>",
        "X-TencentCloud-SecretKey": "<your-secret-key>"
      }
    }
  }
}
  1. Obtain and replace placeholders with your Tencent Cloud credentials and environment ID.

Optional tooling:

  • You can install and use the CloudBase CLI for a smoother experience across IDEs: npm install @cloudbase/cli@latest -g, then run tcb ai. Ensure you follow the CloudBase docs for authentication and usage.

Additional notes

Tips and common issues:

  • For local mode, ensure your environment can access the network to fetch MCP packages and reach Tencent Cloud services during deployment.
  • When using hosted mode, securely store and load your env_id and secret credentials; never hard-code them in shared configurations.
  • If you encounter plugin-related errors in hosted mode, you can disable specific plugins by adjusting the URL query parameters as shown in the Quick Start section. Valid plugin names include env, database, functions, hosting, storage, setup, interactive, rag, cloudrun, gateway, download, security-rule, invite-code, and capi.
  • The MCP server can deploy web apps, small programs, and back-end services; verify the deployment results and ensure the returned access links are reachable from your IDE.
  • Review CloudBase documentation for any region-specific constraints or API changes that may affect MCP deployments.

Related MCP Servers

Sponsor this space

Reach thousands of developers