GEO Analyzer

GEO Analyzer MCP Server enables local content analysis for AI citation signals. You run it as a local MCP endpoint via a lightweight CLI process, and it exposes analysis capabilities that you can call from MCP clients to evaluate content for claim density, information density, sentence structure, and other AI-focused signals.

How to use

You integrate GEO Analyzer into an MCP workflow by configuring your MCP client to connect to the local MCP server as described in your client’s setup. The server analyzes content to estimate signals that influence AI citation and provides detailed scores, key metrics, and recommendations that you can apply to your content planning and editing process.

How to install

Prerequisites you need before installation.

- Node.js 20+
- Anthropic API key (from console.anthropik.com)

Add GEO Analyzer to your MCP configuration by defining the MCP server entry shown here. This config runs the analyzer locally using an executable CLI wrapper.

{
  "mcpServers": {
    "geo-analyzer": {
      "command": "npx",
      "args": ["-y", "@houtini/geo-analyzer@latest"],
      "env": {
        "ANTHROPIC_API_KEY": "sk-ant-..."
      }
    }
  }
}

Troubleshooting

If you encounter issues, check and ensure the API key is correctly provided in the environment config. If your client reports module errors after config changes, restart the client application completely. For content-related errors, confirm you provide enough text (minimum 500 characters) for meaningful analysis.

Migration notes

When upgrading from earlier setups, migrate to the single-token API key approach. Remove any old external service URLs and rely on the local CLI with the ANTHROPIC_API_KEY in the environment to simplify operation.

Development

If you want to build from source during development, follow these steps.

git clone https://github.com/houtini-ai/geo-analyzer.git
cd geo-analyzer
npm install
npm run build

Available tools

analyze_url

Fetches and analyzes published web pages to produce detailed signals and scores.

analyze_text

Analyzes pasted content to extract signals and provide scores and recommendations.