Technical transparency · v2.0

The server delivers the application, not your documents.

The complete document workflow runs in your browser, including output validation and ZIP creation.

01 / 05

1. Opening and validation

A dedicated worker checks the real file signature, PDF structure, encryption state, page count, and adaptive limits without an upload. Password-protected and unreadable files are refused.

02 / 05

2. Previewing

A same-origin copy of PDF.js renders only visible pages plus a small margin. JavaScript evaluation, XFA, and interactive annotation layers are disabled. Temporary canvases and object URLs are released.

03 / 05

3. Editing

The workspace stores page references and operation differences, not copies of entire documents. Original File objects are never modified. Undo and Redo keep at most 50 operations.

04 / 05

4. Export and validation

Workers copy pages, apply supported edits, serialize PDFs, and build ZIP archives locally. A generated PDF is reopened and its page count is checked before a download action is exposed.

05 / 05

5. Offline boundary

After complete caching, the app shell and document engines work offline. The first visit does not claim offline readiness until the Service Worker is installed. Document data is never added to that cache.