Geolook

GeoLook is an open source platform for generative engine optimization, the practice of getting AI answer engines to mention and cite a brand. It runs the whole cycle in one place: measuring current standing, diagnosing why a site is or is not being quoted, setting a strategy, issuing implementation tickets, tracking execution and verifying the result afterward. Monitoring covers seventeen engines, ten sampled automatically through APIs and seven collected manually with a Chrome assistant, among them ChatGPT, Perplexity, Gemini, DeepSeek, Google AI Overviews, Doubao and Metaso, so both Western and Chinese engines are represented. It reports mention rate, rank and citation share, replays the raw answers, and compares competitors against a question bank that can be widened with terms mined from Google autocomplete and Baidu suggestions. The diagnostic side audits the site itself: parsing robots rules, probing firewall and content delivery network behavior using AI crawler user agents, validating llms.txt and scoring passages for how quotable they are. Findings arrive as structured tickets with risk grades, supported by a content workbench and generated assets including llms.txt files, structured data markup and snippets, with publishing confirmed by hand. Client ready reports, scheduled repeat sampling and support for several brands are included. The project is free. It is released under the MIT license and runs on your own machine: the site shows the whole setup as cloning the repository, installing a few Python packages and starting a local interface on a loopback address. No plans or paid tiers exist, and the only running cost is whatever the engine APIs charge for sampling, which can be avoided entirely by collecting answers manually.