GEO Daemon: Track Your Visibility in AI Search Engines

GEO Daemon is an open-source tool developed by Evan Huang, designed to address the pain points of brand visibility monitoring in the AI search era. It automatically queries five major AI search engines including ChatGPT, Perplexity, and Claude weekly to track mentions of specific individuals or brands, generating time-series data and visual dashboards. The tool is written in TypeScript and runs fully automatically via GitHub Actions without manual intervention.

With the rise of AI search engines like ChatGPT and Perplexity, the way users access information has fundamentally changed. Traditional SEO focuses on web page rankings on Google, while GEO (Generative Engine Optimization) focuses on whether content/brands are mentioned or cited by AI models. AI search responses are dynamically generated without traditional result lists, so brands need new ways to monitor their visibility in the AI ecosystem—thus GEO Daemon was born.

GEO Daemon is a personalized generative engine optimization daemon whose core mission is to automatically query multiple mainstream AI search engines weekly, detect mentions of target individuals, and aggregate data into a public dashboard. The tool is written in TypeScript and executes automatically every Sunday at UTC 12:00 via GitHub Actions—once configured, no manual intervention is needed.

The project uses a modular design, with each AI search engine having an independent Provider implementation: Perplexity (real-time search Q&A), OpenAI (ChatGPT Search), Anthropic (Claude), Gemini (Google Assistant), and SerpAPI (traditional search comparison). To add a new engine, simply implement a new Provider class in the src/providers/ directory, which handles API calls, response parsing, and mention detection.

GEO Daemon's workflow: 1. Users define tracking queries in queries.yaml; 2. Iterate through Providers to send queries; 3. Parse responses to detect mentions and their sources; 4. Store results as JSON in data/runs/ (with timestamps); 5. Reconstruct index.json to aggregate historical data; 6. Notify and update the public dashboard via GitHub Actions.

The generated time-series data can help users: track changes in mention frequency across different AI engines; compare differences in model responses; evaluate the effectiveness of content strategies; identify sources of AI citations. This data is of great value for personal brand building, corporate reputation management, and content strategy optimization.

Deployment steps: 1. Fork the repository; 2. Configure API keys in GitHub Secrets; 3. Customize queries.yaml; 4. Enable scheduled triggering of Actions. For local testing, use npm run daemon -- --dry-run. Future outlook: features like fine-grained sentiment analysis, cross-language tracking, deep CMS integration, and industry benchmark comparisons.