Private documents, explicit access and visible limits.
PDFLP separates original files from public page assets, uses short-lived download tokens and gives workspace owners control over search visibility. This page describes the current product design, not certifications we do not hold.
By PDFLP · Last updated
On this page5 sections+
Access control
- Workspace routes require an authenticated database session.
- Session cookies are HttpOnly and SameSite=Lax.
- Passwords use salted scrypt hashes and are never stored in plaintext.
- Email verification is required before workspace login.
- Password resets use single-use, expiring tokens and revoke old sessions.
Document storage
Original PDFs and generated page images are stored in a private Cloudflare R2 bucket accessed through a Worker binding. The public page uses image responses that are checked against the landing page publication state.
The original file requires a valid download token. Tokens expire, have a maximum use count and are stored as hashes.
PDF processing
Rendering runs in an isolated Cloudflare Container. The worker validates the source, renders page images, extracts available text and writes only the necessary outputs to private storage.
Encrypted, corrupt and over-limit PDFs can be rejected. The service does not claim that every uploaded file can be processed.
Data and processors
The service uses Cloudflare for hosting, D1, R2, Queues, Containers, Turnstile and email delivery. Creem is the intended Merchant of Record when paid checkout is enabled.
Read the privacy notice and cookies page for the current data purposes and controls.
Reporting a concern
Send security or privacy questions through the contact form. Include the affected URL and enough detail to reproduce the issue without sending unnecessary sensitive information.
No system is perfectly secure. Avoid uploading confidential documents until you have confirmed that the current product controls meet your organization's requirements.