Skimlane

Personalized reading summaries for webpages and PDFs using on-device, local, or bring-your-own-key AI.

Skimlane is a private AI reading sidekick for Chrome that turns browsing into a custom reading workflow. AI-shaped content needs AI-shaped reading. More of the web is long, repetitive, SEO-shaped, AI-polished, or fully AI-generated. Skimlane gives you a dedicated AI assistant for reading: one that can process pages as you browse, follow your chosen recipes, and turn each source into the structured view you need. Open Skimlane in the Chrome side panel or floating panel, pick a reading recipe, and press Skim. Or turn on Auto Skim and let Skimlane process new pages as you move through the web. You can exclude sites, skim selected text, reopen history, bookmark useful skims, and ask follow-up questions about the current page. Unlike a generic summarizer, Skimlane is built for power users and professional readers. You choose the sections, order, prompts, model, detail level, tone, language, and reading profile. The same page can become a quick digest, product brief, engineering brief, research note, market brief, signal cleaner, knowledge note, share-angle list, or your own custom recipe. Why install Skimlane: - Auto-skim pages as you browse. - Read long pages faster without losing the point. - Decide what is worth reading, saving, sharing, or skipping. - Ask follow-up questions about the current page. - Customize recipes, sections, prompts, tone, language, and reading profile. - Build, import, export, and share reading recipes. - Save useful outputs as structured Markdown. - Use local models where available or bring your own API key. - Keep the workflow in Chrome with no Skimlane-owned backend. Key features: - Auto Skim: process new pages automatically with your default recipe and model. - Site exclusions: prevent Auto Skim from running on domains you choose. - Selection-first reading: skim only selected text when you want focus. - Quick Digest: TL;DR, source signal, main points, and follow-up questions. - Worth Reading?: a read, skip, or save judgment with reasons. - Full Skimlane: a deeper structured skim for dense sources. - Signal Cleaner: unsupported claims, repeated fluff, generic AI signal, and the actual bottom line. - Knowledge Note: reusable notes with concepts, quotes, tags, and links to explore. - PM Brief: user impact, product implications, roadmap relevance, risks, and open questions. - Engineering Brief: technical takeaways, implementation notes, tradeoffs, and jargon. - Research Note: thesis, method, evidence, limitations, and unsupported claims. - Market Brief: market signal, customer pain, competitor angle, and distribution lessons. - Share Angles: quotes, reader themes, best replies, and ways to share the source. - Follow-up chat: ask questions about the current page after a skim. - History and bookmarks: reopen past skims and keep the useful ones. Automation that stays under your control: Skimlane can stay active beside you while you browse. Auto Skim runs on new pages with your chosen default recipe and source. You can turn it off anytime, exclude any website, or use Skimlane only when you explicitly press Skim. Custom output for power users: A normal AI summary treats every reader the same. Skimlane lets you build a reusable way to read. Choose the sections you want, reorder them, edit prompts, create custom sections, export recipes, and import them wherever you use Skimlane. Personalized reading: Skimlane can use your reading profile to tailor outputs. Add your background, what matters to you, what to exclude, preferred output language, and tone. The result is less generic and more aligned with how you actually consume content. Model choice: Use an on-device model where available, connect a local model, or bring your own API key for a supported hosted model provider. Skimlane lets you choose the model source that fits your privacy, cost, and quality preferences. Privacy and control: Skimlane is local-first. Settings, recipes, prompts, profile, API keys, history, saved skims, and chats are stored in Chrome extension storage. There is no Skimlane account, backend, tracking, or cloud knowledge base. If you choose a hosted model, your browser sends the requested content directly to that provider. You can clear history, saved skims, snapshots, chats, settings, recipes, prompts, and API keys from Data & Privacy. Useful for: - Reviewing technical and market content. - Scanning documentation, RFCs, architecture posts, and issue threads. - Identifying claims, evidence, assumptions, and gaps. - Tracking market signals and competitor perspectives. - Turning source material into reusable explanations. - Saving structured Markdown to your notes. - Configuring an AI-assisted reading workflow.