Media
4 items
Description
Detect and categorize website dependencies for internet filter management Dependency Analyzer captures the third-party domains a website talks to and sorts them into clear, actionable categories — so you can build internet-filter allowlists, audit a site's dependencies, or just understand what a page is really loading. Click Start Capture, browse the site normally, then Analyze with AI. Every domain is sorted into one of four categories with a plain-English note on what it does: • First Party — the site's own domain. • Specific Dependencies — services this site needs that you'd add to its allowlist (its CDN, APIs, integrated vendors). • Generic Dependencies — ubiquitous providers (Google Fonts, jsDelivr, Stripe, …) already allowed almost everywhere, so they don't need per-site whitelisting. • Noise — ads, trackers, analytics, and other optional services. FEATURES • Live capture of every domain a tab contacts, with request counts and the exact request paths. • AI classification that reads what each domain actually serves on the page — a content CDN is treated differently from an ad pixel, even on the same domain. • One-click "Copy whitelist" of the domains you actually need to allow. • Subdomain mode to classify each subdomain individually, shown as a tidy tree. • Manual mode — paste a list of hostnames or URLs to categorize without capturing. • Results cached locally so you're not re-billed for sites you've already analyzed. • Light / dark / system theme and adjustable font size. REQUIREMENTS — BRING YOUR OWN API KEY Classification uses Anthropic's Claude (the low-cost Haiku model) with YOUR OWN API key, which you add in Options. A free Anthropic account plus a small amount of credit is all you need — analyzing a site typically costs a fraction of a cent, and re-analyzing a cached site is free. The extension is fully usable for capturing domains without a key; the key is only needed for AI categorization. PRIVACY There is no developer-operated server and no analytics or tracking of any kind. When you run an analysis, domain names plus a sample of request paths (query strings removed) are sent to the Anthropic API using your key, and domain names are sent to DuckDuckGo for short descriptions. The extension also fetches each analyzed domain's public homepage to read its title/description. Your API key, cache, and settings stay in your browser. Full details: https://github.com/aldomanpi/DepAnalyzer/blob/main/PRIVACY.md WHO IT'S FOR Network and content-filter administrators building allowlists, security and privacy researchers, and web developers auditing third-party dependencies.
rating_count is the Chrome Web Store ratings count, not a written-review count.