Phil's Web Tools

Phil's Web Tools puts the page-inspection and browsing-fix tools you actually use into one calm Chrome side panel. No server, no telemetry, no data leaving your browser. CAPTURE - Element picker plus Main-article and Selected-text extraction - Reliable Markdown (Readability + Turndown) and clean semantic HTML - Table picker with Markdown, CSV, and JSON output (handles colspan/rowspan) - Form-field picker with labels, current values, and dropdown options - Content-link collector that strips nav, breadcrumbs, and duplicates AUDIT - Basic SEO: metadata, headings, alt text, links, Open Graph, JSON-LD - Search + social-card previews - JSON-LD validation with Schema.org type detection - robots.txt, sitemap, and feed discovery - Loaded-asset inspector (images, fonts, styles, scripts) with sizes - Page-wide design palette and CSS-variable export - Local, signature-based technology detector with visible evidence INSPECT & FIX - Cookie and localStorage viewer/editor for the current site - Per-tab redirect chain with status codes - Global right-click and text-selection recovery on pages that block them - JSON formatter, minifier, and automatic response viewer - Tab tools: dedupe, search, group-by-site, and bulk close - Site helpers for Google Search/Images and Gmail ON-DEVICE AI (optional, off by default) - Summarise the page, ask questions about it, and explain any selection - AI SEO suggestions and topic-based tab grouping - Quiet inline enrichment of your cookie, link, tech, and redirect results - Runs entirely in your browser on WebGPU — no server, no API key, no page content ever sent anywhere. Turn it on in Settings; a language model downloads once from the Hugging Face hub and is cached locally, then works offline. Everything runs locally. Phil's Web Tools has no accounts, no analytics, and loads no remote code. (The optional AI feature downloads model weights — data, not code — one time; nothing you view is ever transmitted.)