Integrations

The kluster.ai Code verification service is designed to integrate directly into your IDE workflow, providing real-time code analysis as you develop. By leveraging MCP, Code verification works seamlessly with AI coding assistants to catch issues before they reach your codebase.

For Cursor users, a one-click installation process is available that handles all setup automatically. See the Code Quick Start guide for the fastest way to get started.

Prerequisites

Before getting started, ensure you have:

• A kluster.ai account: Sign up on the kluster.ai platform if you don't have one.
• A kluster.ai API key: After signing in, go to the API Keys section and create a new key. For detailed instructions, check out the Get an API key guide.

Supported IDEs

Code verification works with any MCP-compatible client, including:

• Cursor: One-click installation with automatic MCP server setup (most popular).
• Windsurf: AI coding assistant with Cascade and MCP integration.
• VS Code: Manual MCP configuration via .vscode/mcp.json.
• Kilo Code: AI coding assistant with streamlined MCP configuration.
• Claude Code: Manual MCP configuration via .claude/mcp.json.
• Cline: Open-source AI coding agent for VS Code with MCP marketplace.
• Roo Code: AI-powered dev team extension with dual configuration support.
• Any MCP-compatible IDE: Manual configuration using the MCP server details below.

MCP configuration

Add the following to your MCP configuration file:

{  
    "mcpServers": {  
        "Kluster-Verify-Code-MCP": {  
            "command": "npx",  
            "args": [  
                "@klusterai/kluster-verify-code-mcp@latest"  
            ],  
            "env": {  
                "KLUSTER_API_KEY": "your-api-key-here"  
            }  
        }  
    }  
}  

Setup instructions

CursorWindsurfVS CodeKilo CodeClaude CodeClineRoo CodeOther MCP Clients

One-click installation (recommended):

For the fastest setup, use our one-click installation process described in the Get started with Verify Code guide.

Manual configuration:

1. Open Cursor Settings by clicking the gear icon on the top right corner.

2. Navigate to Tools & Integrations in the left sidebar.

3. Under MCP Tools, click Add Custom MCP.

   

4. This opens the mcp.json configuration file. Add the Kluster-Verify-Code-MCP server configuration:

   

5. Save the configuration file and return to Cursor Settings.

The Kluster-Verify-Code-MCP server will now appear with both tools enabled:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

1. Open Settings by clicking the gear icon or using the command palette.

2. Click Windsurf Settings.

   

3. Navigate to Cascade in the left sidebar.

4. Select Manage MCPs to access the MCP configuration.

   

5. Click View raw config to access the MCP configuration file.

   

6. Add the Kluster-Verify-Code-MCP configuration to your mcp_config.json:

   

7. Save the configuration and refresh.

The Kluster-Verify-Code-MCP will appear with both tools enabled:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

1. Create a .vscode folder in your project root if it doesn't exist.

2. Create a mcp.json file inside the .vscode folder.

3. Add the following configuration:

{
  "servers": {
    "Kluster-Verify-Code": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@klusterai/kluster-verify-code-mcp@latest"
      ],
      "env": {
        "KLUSTER_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Once mcp.json is saved, to verify the successful installation of the server:

1. Click on the tools icon in the bottom right corner.

2. Select the Kluster-Verify-Code MCP server from the list.

The Kluster-Verify-Code MCP Server will provide both tools:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

1. Open Settings and navigate to MCP Servers.

2. Select the Installed tab.

3. Click Edit Project MCP to open the MCP configuration.

4. Paste the content of the MCP config shown above with your API key.

Then you should see the installed Kluster-Verify-Code-MCP server with both tools enabled:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

1. Create or edit .claude/mcp.json in your project.
2. Add the Code MCP server configuration shown above with your API key.
3. Restart Claude Code and the tools will be available immediately.
4. Run the command /mcp to check the status of the tools.

1. Open Cline and click MCP Servers in the left sidebar.

2. Select the Installed tab.

3. Click on Configure MCP Servers, which will open the cline_mcp_settings.json configuration file. Add the Kluster-Verify-Code-MCP server configuration.

   

4. Save the configuration.

The Kluster-Verify-Code-MCP server will now appear with both tools enabled:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

1. Open Roo Code and click on MCP Servers in the left sidebar.

2. Click Edit Project MCP to open the MCP configuration for your project.

   

3. The configuration file will open at .roo/mcp_settings.json. Add the Kluster-Verify-Code-MCP server configuration shown above.

4. Save the file.

The Kluster-Verify-Code-MCP server will appear with both tools enabled:

• kluster_code_review_auto: For code security and quality verification.
• kluster_dependency_validator: For dependency validation.

For any other MCP-compatible IDE or client:

1. Locate your MCP configuration file (varies by client).
2. Add the Code MCP server configuration shown above.
3. Restart your IDE if required by the client.
4. The tools should now be available in your AI assistant.

Available tools

For detailed information about each tool, see our Tools reference.