- Description
- Screenshots
- Features
- Requirements
- Installation
- Usage
- Configuration
- Permissions
- Limitations
- Architecture
- Credits
- Contributing
- License
A Manifest V3 Chrome extension that saves the page you are looking at in up to five formats with one click: a self-contained HTML file, clean Markdown, a full-page PNG, a screenshot PDF and a print PDF with selectable text. By default the files arrive as one ZIP in your Downloads folder.
Everything happens inside your browser. Nothing is sent to a server.
The popup: pick any combination of the five output formats, then Archive.
| Format | What it captures | How |
|---|---|---|
| HTML | The whole page as one self-contained file, with stylesheets and images inlined and everything that could run code (scripts, event handlers, javascript:/data:text/html URLs, iframes, embeds, meta refreshes) stripped, plus a restrictive CSP so the file stays inert if opened from disk |
A custom sanitizing serializer over the live DOM |
| Markdown | Just the article text, with YAML front matter | Readability + Turndown |
| PNG | A full-page screenshot, exactly as the browser rendered it | Scrolls the page, captures each screen and stitches them together |
| Screenshot PDF | The full-page screenshot, split across A4 pages | jsPDF |
| Print PDF | The page as Chrome prints it, with selectable text | Chrome's own Page.printToPDF |
- One click, any combination of the five formats
- One ZIP or separate files, your choice
- A configurable file name built from the date, site, page title and a timestamp
- A keyboard shortcut that archives with your saved default formats
- Markdown that falls back gracefully: when a page has no recognisable article, the whole body is converted instead
- Local only: no account, no server, no tracking
- Google Chrome 109 or newer, or another Chromium-based browser (Edge, Brave, Arc) that supports Manifest V3 offscreen documents
- Developer mode turned on at
chrome://extensions, to load the extension unpacked - Node.js 20+ and npm, only if you want to update the bundled libraries or
run the tests. The built libraries are already in
vendor/.
Or skip the clone: download the zip from the latest release, unzip it and load that folder.
git clone https://github.com/geoffmyers/webpage-archiver.gitThen load it into Chrome:
- Open
chrome://extensions. - Turn on Developer mode (top right).
- Click Load unpacked and choose the
webpage-archiverfolder. - Pin Webpage Archiver to the toolbar from the extensions menu.
To rebuild the bundled libraries after changing a dependency:
cd webpage-archiver
npm install
npm run build # copies the libraries from node_modules into vendor/- Open the page you want to keep.
- Click the Webpage Archiver icon in the toolbar.
- Tick the formats you want: HTML, Markdown, PNG, Screenshot PDF and Print PDF are all on by default.
- Click Archive. A progress bar shows each step, then a list of what was saved.
The PNG and screenshot PDF are made by scrolling the page, so leave the tab in front until the progress bar finishes.
Ctrl+Shift+S (⌘+Shift+S on macOS) archives the current tab with the
formats saved in the options. Change the key at chrome://extensions/shortcuts.
The default pattern is {date}_{hostname}_{title}. Spaces become hyphens,
characters that are not allowed in file names are removed, and names are cut to
120 characters. For a page titled "Article Title" on www.example.com:
2026-02-17_example.com_Article-Title.zip ← default: one ZIP containing
2026-02-17_example.com_Article-Title.html
2026-02-17_example.com_Article-Title.md
2026-02-17_example.com_Article-Title.png
2026-02-17_example.com_Article-Title.screenshot.pdf
2026-02-17_example.com_Article-Title.print.pdf
When only one of the two PDF formats is selected, its file is just .pdf.
{date} is the UTC date, and a leading www. is dropped from {hostname}.
The Markdown format keeps only the main article, without navigation, ads, sidebars or footers, and starts with front matter. Fields the page does not provide are left out.
---
title: "Article Title"
url: "https://example.com/article"
archived: "2026-02-17T00:00:00.000Z"
author: "Author Name"
excerpt: "A brief summary..."
siteName: "Example.com"
---
Clean article content here...Open the options from the Options link in the popup, or from
chrome://extensions → Webpage Archiver → Details → Extension options.
| Setting | Default | What it does |
|---|---|---|
| Default formats | All five | Which formats are ticked when the popup opens, and which the shortcut uses |
| Filename pattern | {date}_{hostname}_{title} |
Tokens: {date}, {hostname}, {title}, {timestamp} (milliseconds since 1970) |
| Subfolder | (none) | Saves into a folder inside Downloads |
| Bundle all formats into a single ZIP file | On | Off saves each format as a separate download |
Settings are stored with chrome.storage.sync, so they follow your Chrome
profile.
| Permission | Why |
|---|---|
<all_urls> (a host permission, not listed under permissions) |
Read the page you asked to archive |
scripting |
Run the capture script in that page |
downloads |
Save the files |
storage |
Remember your settings |
offscreen |
Stitch screenshots, build the PDF and pack the ZIP; a Manifest V3 service worker has no DOM to do it in |
debugger |
Produce the print PDF through Chrome's Page.printToPDF. Chrome shows a "started debugging this browser" bar while it runs |
- Pages without a clear article (dashboards, single-page apps, social feeds) give Readability nothing to extract, so the Markdown falls back to the whole page body and is noisier.
- Cross-origin images cannot be inlined into the HTML archive because the browser blocks reading them. They keep their original URLs.
- Very long pages are cut off at 32,000 pixels in the PNG and screenshot PDF. The results list says so when it happens.
- Restricted pages such as
chrome://pages and the Chrome Web Store cannot be captured, because Chrome does not let extensions run scripts there.
popup.js ──► service-worker.js ──► content-script.js (in the page)
│ HTML serialisation, Readability, Turndown
│
├──► captureVisibleTab, screen by screen ──► offscreen.js
│ stitch PNG, jsPDF, JSZip
├──► chrome.debugger: Page.printToPDF
└──► chrome.downloads
| Path | Role |
|---|---|
manifest.json |
Permissions, service worker, popup, options page and shortcut |
src/popup/ |
Format picker, progress and results |
src/background/service-worker.js |
Orchestrates a capture: injects the content script, scrolls and screenshots, builds file names, downloads |
src/content/content-script.js |
Runs in the page: orchestrates capture, Markdown extraction |
src/content/html-sanitizer.js |
Runs in the page: builds the sanitized single-file HTML archive (no chrome.* calls, so it can also be loaded directly by tests) |
src/offscreen/ |
DOM work the service worker cannot do: stitching, PDF and ZIP |
src/options/ |
Settings page |
vendor/ |
Bundled third-party libraries, committed so the extension loads with no build step |
build.js |
Copies those libraries out of node_modules |
tests/ |
Playwright tests |
See ARCHITECTURE.md for more detail.
| Library | License | Used for |
|---|---|---|
| @mozilla/readability | Apache-2.0 | Finding the article in a page |
| Turndown and turndown-plugin-gfm | MIT | HTML to Markdown, including tables |
| jsPDF | MIT | The screenshot PDF |
| JSZip | MIT or GPL-3.0 | The ZIP bundle |
| Playwright | Apache-2.0 | Tests |
The icon, in the extension and here, is the Font Awesome
box-archive glyph, used under CC BY 4.0.
Chrome is a trademark of Google LLC. This extension is not affiliated with or endorsed by Google.
Written by Geoff Myers.
Bug reports and pull requests are welcome. See CONTRIBUTING.md for setup, checks and how this repository is published.
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See LICENSE.md for the full text of the GNU General Public License.
SPDX-License-Identifier: GPL-3.0-or-later