Documentation

Projects and data

This page explains how uploaded files, access tokens, and project data are managed.

Input formats

PDF

Upload a compiled paper directly as a PDF. CScope extracts the manuscript structure, references, figures, and tables, then uses them in the later analysis stages.

LaTeX ZIP

Upload the complete LaTeX archive and select its main .tex file. LaTeX projects support source export and, when local compilation is configured on the server, can build a compiled PDF.

Access token

Each project receives an access token beginning with ppr_. The complete token is shown only once and must be supplied whenever the workspace is opened again.

Store the token carefully.Anyone who has it can access the corresponding project. Do not include it in public screenshots, issues, chat messages, or shared documents.

Retention

By default, a project is deleted after 30 consecutive days without access. Opening the workspace refreshes its last-access time. Deleting a project removes the original upload, parsed files, analysis results, and project logs.

What a project stores

The uploaded PDF or LaTeX archive.
Parsed paper text, section structure, figures, and tables.
Search results and the output of each check.
Project-level model interaction records used to investigate output failures.

These files stay on the server running CScope and are not part of the source repository.

How data is used

Paper content is sent to the model provider selected for the project. When searching for related work, CScope also sends topic-based queries to the configured search services. The exact data boundary depends on the providers you use. Read their privacy terms before submitting sensitive or unpublished work.