Atomic Clipper

Clip any web content into a library inside Chrome. Export to Markdown for NotebookLM, Perplexity, or any LLM.

Atomic Clipper Clip any part of any web page into a library inside Chrome, then export it as Markdown for NotebookLM, Perplexity, or any LLM. How it works Click the toolbar icon, then Start Clipping. A crosshair appears. Select what you want, and a save panel opens where you assign a category. Press Esc at any point and Atomic Clipper leaves the page as it found it. Five ways to capture Point and click. Hover to highlight any element, then click it. Drag across a passage. When what you want does not line up with a single element, select it by dragging. The clip keeps the headings and lists inside your selection instead of flattening them. Hold Ctrl or Cmd to collect several regions. Keep the key down and click or drag as many parts of the page as you like. A tray counts them, and Remove last undoes a mistake. Release the key and they save as one clip, divided by horizontal rules. Clip an image on its own. A figure, chart or photograph with no text around it is still a clip. The save panel shows you a thumbnail before you save. Let Atomic Clipper find the article. On a long page, click Clip Article instead. Atomic Clipper picks the main content block and opens the save panel. Your library Every clip lands in your library, grouped by category. Open it from the popup footer. Copy MD copies one clip as a standalone note, YAML frontmatter already on it: title, source URL, tags, and the date you saved it. Paste it straight into Obsidian or any Markdown vault. Export .md takes a whole category as one file, frontmatter at the top and every clip below. Why Markdown Markdown keeps your titles, headings, paragraphs and quotes intact, in plain text. No proprietary format, no lock-in. The files open in any note-taking app or text editor. If an image in a clip no longer loads, the export marks it in place, so you never inherit a broken link with no explanation. Privacy and permissions Everything stays on your device. Atomic Clipper uses Chrome's local storage: no account, no cloud sync, no backend. Delete one clip, or all of them, anytime, from the library. Atomic Clipper makes no network requests of its own. It sends no analytics, no crash reports, no usage data. Nobody behind this extension receives anything about you or about what you clip. Your browser does make ordinary image requests to third-party hosts, in three situations: - the library shows the images in a clip - the save panel previews an image clip before you save it - at capture time, Atomic Clipper tests whether each image still loads, so the export can mark the ones that fail All three ask for the image at its original address, the one the page you clipped already used. Atomic Clipper sends no referrer, so the host learns nothing about the page or the extension. The requests carry no clip content, no category, and no other stored data. Nothing reaches Atomic Clipper or anyone connected with it. What the extension stores for an image: - a linked image uses an http or https address. Atomic Clipper saves the address and never downloads the file - an embedded image uses a data: address, which holds the image data itself. Atomic Clipper saves that data with the clip. Above roughly 100 KB, the save panel asks you to keep or skip it Three permissions, nothing else: - activeTab — works on the current tab, only when you click the icon - scripting — injects the element picker into the page - storage — saves your clips locally No host permissions. No access to your browsing history or your cookies. Open source MIT license. The picker, extractor, storage and export logic are all on GitHub, open to audit. Known limitations - Atomic Clipper cannot activate on browser system pages (chrome://, about:, extension pages) - It may not clip correctly behind a login wall or paywall