Mapping Website Internal Links

A Python crawler that starts from a chosen address, visits up to a set number of internal pages and collects SEO, accessibility, performance and link structure data for each one. The results are drawn as a D3 network graph in a local browser app, where pages appear as nodes and internal links as connections, with a scorecard for every selected page. A Flask backend sends each page's structured crawl data to Claude, which reviews SEO and accessibility and suggests improvements. Setup needs a Python environment, an Anthropic key in a local settings file and a page limit for the crawl. It is useful for seeing how the internal linking of a domain hangs together before changing navigation or content. Released under the MIT license, so it is free to use.