Shipped today — v0.1.0

Viewer with thumbnails, zoom, fit modes and keyboard navigation
Page operations: reorder by dragging, rotate, delete, duplicate, extract, merge, split
Annotation: text boxes, highlights, freehand ink, shapes, images, drawn signatures
Fill and flatten existing AcroForm form fields
Stamp Studio — layouts, frames, inks, distress, variable fields, PNG export, .stamps packs
Workflows — record, edit, save as a file, run on one document or a folder
Bates numbering, page numbers, headers, footers, watermarks, metadata removal
Batch processing: folder in, ZIP out, entirely on your machine
Local Processing Monitor and Safe Mode, with every export reopened and verified
Offline support and a single-file portable build

Not shipped, and honestly labelled

Permanent redaction

This is the most requested thing and it is not here. Nearly every free tool that claims redaction draws a black rectangle over the text and leaves the text underneath, selectable by anyone who copies and pastes. That failure has exposed witness names and settlement figures in real cases.

Doing it properly means removing the content from the page's content stream, sweeping the associated metadata, annotations and hidden text, then re-running text extraction on the export to prove the terms are gone — and showing you that proof. That is real work and we would rather ship it late than ship a rectangle. Until then the answer is no, and the tool will not offer you anything that looks like a yes.

Optical character recognition

Making scanned documents searchable, locally, using a WASM OCR engine. Realistic, and large — the model files are bigger than the entire editor, so it needs to be optional and pausable rather than bundled.

PDF X-ray

Showing what a PDF actually contains: embedded fonts, true image resolution, hidden layers, JavaScript, attached files, metadata and producer history. Quietly a security tool — no free viewer tells you that the invoice you were sent contains a script and an embedded file.

Compare two versions

Visual and text difference between two PDFs.

Decided against

Not building	Why
Our own PDF parser and writer	PDF.js and pdf-lib are mature and battle-tested. Rewriting them is a multi-year detour that would sink the project and produce something worse.
Editing existing body text	Reflowing text in an existing PDF corrupts layout in ways that are hard to spot. Even the most expensive desktop editors mangle it. Silent damage to a contract destroys trust permanently.
XFA forms	Deprecated by the format's own originator. Real engineering cost, shrinking user base.
Cryptographic digital signatures	Certificate handling in a browser is a support burden out of proportion to the need. Visual signatures cover the common case and we label them honestly.
Accounts, sync and cloud storage	Each one contradicts the entire premise.
AI features needing an API key	Breaks "no keys, no accounts" and introduces a per-user cost we would eventually have to charge for.
A native mobile app	The web app is the app. It installs to your home screen and works offline already.

How priorities get set

By what people actually hit, not by what sounds impressive. If a PDF fails to open, that outranks any new feature on this page. Tell us what broke — awkward files are how this gets better, and they end up in the test corpus so the same thing cannot break twice.

Roadmap