pagecap

About

pagecap captures any web page at the fidelity the page deserves — vector PDF,
high-DPR raster (PNG / JPEG / WebP / AVIF), or standalone HTML — and saves it
to your machine. Nothing leaves your device. Zero telemetry, zero remote
calls, zero analytics, open-source MIT.

WHY PAGECAP

Existing full-page screenshot tools fail on at least one of four axes:

• Resolution at zoom. 1× raster captures alias the moment you zoom past 200% —
  text becomes a smudge. pagecap’s default is vector PDF: text stays as glyphs,
  infinitely zoomable.
• File size. A high-DPR PNG of a long article can be 50–200 MB. pagecap’s
  vector PDF of the same article is sub-MB; AVIF for image-heavy pages is a
  fraction of equivalent PNG.
• Dynamic content. Lazy images, virtual scrollers, sticky headers, cookie
  banners. pagecap runs a page-prep handler pipeline before capture (pause
  animations, prime lazy images, exhaust infinite scroll, walk virtual
  scrollers, neutralize sticky/fixed elements, optional cookie-banner dismiss).
• Format flexibility. Most tools ship PNG only. pagecap ships PDF, PNG, JPEG,
  WebP, AVIF, and standalone HTML — pick the right tool for the page.

WHEN TO USE WHICH FORMAT

• Text-heavy article, README, docs → PDF. Vector text → infinite zoom; sub-MB
  even for a 30-screen page.
• Image-heavy galleries / photo grids → AVIF. Smallest format for natural
  imagery at good quality.
• Mixed text + screenshots → PDF. The safe default.
• Needs to re-flow / be responsive → HTML. Inlines same-origin styles and
  images; strips scripts and iframes.
• High-res hero shot → PNG at DPR 2× or 3×. Lossless retina quality.
• Sharing on chat or low bandwidth → JPEG at 92% quality.

HOW IT WORKS

1. Open the page you want to capture.
2. Click the pagecap toolbar icon.
3. The popup shows a recommended format based on a 10 ms read of the page
   (text length, image count, page height). Pick a format and any options.
4. Click “Capture as <format>”. Save the file through Chrome’s normal Save-As
   dialog.

Chrome shows a yellow “DevTools is debugging this tab” banner during PDF and
high-DPR raster captures — that is how pagecap obtains vector / high-fidelity
output via the Chrome DevTools Protocol. The banner disappears when capture
finishes. A first-run tooltip in the popup explains this.

OPTIONS

• Dismiss cookie / consent modals. Best-effort heuristic, off by default.
• Exhaust infinite scroll. Scrolls to the bottom until the page stops growing
  (capped at 1 M CSS px / 1000 iterations). Off by default.
• Always show Save-As dialog. When off, Chrome reuses your last save folder.
• Filename template, default format, default DPR, default JPEG / AVIF quality,
  capture mode (debugger vs always-stitch) — all editable from the Options
  page.

PRIVACY (THE WHOLE STORY)

pagecap makes zero outbound network calls of its own. There is no extension
server. There is no analytics SDK. The only network traffic that exists is
the browser fetching the page you are already on. Your captures, your
settings, and your usage stay on your machine. The full per-permission
rationale is in docs/permissions.md in the repository.

KNOWN LIMITATIONS

• Cross-origin iframes render blank in PDF (Chromium limitation).
• Heavily virtualized feeds (react-window, tanstack-virtual) cannot always be
  fully expanded to a single snapshot. PDF multi-page handles them better
  than raster.
• HTML capture uses a minimal serializer in v1.0; full SingleFile integration
  is a v1.1 target. Pages with extensive shadow DOM or JS-rendered content
  lose dynamic state.
• Cookie-banner dismiss is heuristic. Opt-in. May click the wrong button on
  non-English UIs or unusual layouts.

SOURCE & SUPPORT

Source code, issue tracker, and changelog:
https://github.com/bettyguo/capture-engine

MIT licensed. Contributions welcome.