home / mcp / hubble mcp server

Hubble MCP Server

Provides Hubble data APIs for keyword search, trends, graph relations, and crawling via an MCP interface.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "ascentkorea-hubble_mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "C:\\Users\\XXXX\\Documents\\hubble\\hubble_mcp",
        "run",
        "data_api.py"
      ],
      "env": {
        "HUBBLE_API_KEY": "xxx-xxxx-xxxx-xxx"
      }
    }
  }
}

You run the Hubble MCP Server to access Hubble data APIs through a standardized MCP interface. This setup lets you launch a local MCP instance that exposes data endpoint tools such as keyword analysis, search paths, and Google SERP/Crawl integrations via the MCP protocol.

How to use

You interact with the Hubble MCP Server through an MCP client. Start the server using one of the provided stdio configurations, then connect your MCP client to the local runtime. Common use cases include querying keyword information, exploring keyword relationships, and crawling SERP data or web pages through the same MCP interface.

How to install

Prerequisites you need before starting:
- Python environment (the server is written in Python).
- MCP runtime for stdio execution (uvx-compatible). If you are using a Windows or macOS/Linux setup, install uvx per the platform instructions.

Install the Hubble MCP Server artifacts and prepare the data directory: - Download the Hubble MCP Server release and unzip to a local path. - Obtain a HUBBLE_API_KEY for authenticating your requests.

Configure Claude Desktop to launch the MCP server by adding an MCP server entry that runs uv with the appropriate data_api.py file and directory. You will need to specify the HUBBLE_API_KEY in the environment section.

Configuration and example MCP configs

The following MCP runtime configurations show how to run the Hubble MCP Server using stdio. The examples include platform-specific paths and the environment variable for the API key.

Additional configuration examples

{
  "mcpServers": {
    "hubble_windows": {
      "type": "stdio",
      "name": "hubble_windows",
      "command": "uv",
      "args": [
        "--directory",
        "C:\\Users\\XXXX\\Documents\\hubble\\hubble_mcp",
        "run",
        "data_api.py"
      ],
      "env": {
        "HUBBLE_API_KEY": "xxx-xxxx-xxxx-xxx"
      }
    }
  }
}

MacOS/Linux configuration example

{
  "mcpServers": {
    "hubble_macos": {
      "type": "stdio",
      "name": "hubble_macos",
      "command": "/Users/XXX/.local/bin/uv",
      "args": [
        "--directory",
        "/Users/XXX/workspace/hubble/hubble_mcp",
        "run",
        "data_api.py"
      ],
      "env": {
        "HUBBLE_API_KEY": "xxx-xxxx-xxxx-xxx"
      }
    }
  }
}

Security and keys

Keep your HUBBLE_API_KEY secret and rotate it periodically. Do not share the key in public forums or logs. When running on shared machines, ensure the directory permissions restrict access to the data_api.py and related MCP runtime artifacts.

Troubleshooting notes

If the MCP client reports API failures or unexpected data, verify that HUBBLE_API_KEY is correctly set in the environment and that the data_api.py path points to the actual data service bundled with the Hubble MCP Server release. Ensure the uv runtime is accessible in your system path or specify the absolute path to the uv executable. If issues persist, re-check the data directory and confirm you are using the latest release version for compatibility.

Tools and endpoints available

The Hubble MCP Server exposes endpoints and tools to access and manipulate data, including keyword search metrics, graph relationships, and various crawlers. See the tool list for the complete set of capabilities.

Available tools

get_search_path

API to retrieve search paths for keyword analysis.

get_keyword_info

Provides keyword metrics such as monthly search volume, SERP features, and related intent in Korea.

get_graph_info

Returns information about relationships between keywords.

crawl_google_serp

Crawls Google SERP data for a given keyword.

crawl_web_page

Crawls a general web page to extract content and metadata.

crawl_google_suggest_extension

Crawls Google Suggest for extended keyword ideas.

crawl_google_trends

Accesses Google Trends data for keyword topics.