Installation Guide (JSON only)

Maguyva.ai Documentation

Connect client to MCP server

🗄️

Your Github account must be connected to your Maguyva account at app.maguyva.ai.

🔑

You must obtain a Maguyva personal API key from app.maguyva.ai.

Set your key and repository selection in JSON file

⚠️

Your personal API key will be stored in a JSON file. This means a secret is stored along with application config. You should protect it and ensure it’s not shared.

ℹ️

Replace “mgv_xxxx” in the examples below with your personal Maguyva API key.

ℹ️

Replace “github_user_account/repository_name” with the github name and repo you’re using.

Connect the MCP via mcp-remote

Standard config

{
  "mcpServers": {
    "maguyva": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://maguyva.tools/mcp",
        "--allow-http",
        "--header",
        "Authorization: Bearer mgv_xxxx",
        "--header",
        "Accept: application/json, text/event-stream",
        "--header",
        "Content-Type: application/json"
      ],
      "env": {
        "DEFAULT_REPOSITORY": "github_user_account/repository_name"
      }
    }
  }
}

⚠️

The examples below have not been updated to specify a default repository.

Claude Code

edit your .mcp.json file and paste in the json

Claude Desktop

Claude Desktop, navigate to Claude > Settings… > Developer > Edit Config to open the claude_desktop_config.json

Cursor

Edit your ~/.cursor/mcp.json file and paste in the json

Gemini CLI

# Install MCP server
npm install -g @maguyva/mcp@latest

# Configure in Gemini CLI settings
gemini config set mcp.servers.maguyva "npx @maguyva/mcp@latest"

Goose

# goose.yaml
mcp_servers:
  maguyva:
    command: npx
    args:
      - "@maguyva/mcp@latest"

LM Studio

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

Qodo Gen

{
  "mcpServers": {
    "maguyva": {
      "command": "npx",
      "args": [
        "@maguyva/mcp@latest"
      ]
    }
  }
}

VS Code

{
  "mcp.servers": {
    "maguyva": {
      "command": "npx",
      "args": [
        "@maguyva/mcp@latest"
      ]
    }
  }
}

Windsurf

{
  "mcpServers": {
    "maguyva": {
      "command": "npx",
      "args": [
        "@maguyva/mcp@latest"
      ]
    }
  }
}

Getting Help

Free Trial - Test Maguyva.ai with your codebase
Community Support - Get help from other developers
Contact Us - Direct support for technical questions

Ready to supercharge your AI workflow? Most developers see immediate improvements in their first session.

Start Your Free Trial →

Installation Guide (NPM method)MCP API Reference