R RIGOR Research audit for the social sciences
on the RIGOR server AI models How it works Samples

Audit your manuscript before you submit it.

Findings verified against the paper's own contents, with the repair each one needs.

01 Manuscript

02 Which AI should read it?

Choose your provider and model. RIGOR prices each model against this manuscript, and later offers only the spend limits that can actually run it. It never charges you, never sees your password, and never stores your key.

The key is held in memory by the RIGOR process running your audit: on your own machine when you run RIGOR locally, and on the RIGOR server when you use the hosted site, for this session only. It is never written to disk, never stored in your browser, and never appears in the audit record or the report. Forget key removes it immediately, and closing the session removes it anyway.

  • RIGOR never sees your provider account or its password
  • It cannot read your other conversations or your email
  • Your API key is held for this session only and never written down
  • The copy of your manuscript is deleted when the audit ends

Your manuscript is uploaded to the server running this audit, kept in a private working folder for the length of the run and deleted when it ends. It is then sent to the provider you choose, under your own account, because a model has to read it to audit it.

03 What stage is it at?

Finished paper, or work still in development. This decides whether RIGOR audits what is there or also proposes the smallest work that would make the claims defensible.

04 How should RIGOR audit it?

Audit depth (how deep?)

Focus

Choose up to 3 specific areas. Full paper is not available yet.

05 How much may it spend?

A ceiling, not a price. Your provider bills what the audit actually uses, which for a short paper is often under a dollar; the limit is the most RIGOR is allowed to spend before it must stop and write. A tier below the minimum for this manuscript, model and audit is shown but cannot be chosen: the figure comes from the same check that gates the run, so nothing offered here can be refused at Start.

RIGOR measures the real tokens as it runs and reserves what a report costs, so an admitted budget produces a complete report rather than running out with nothing. A larger limit buys more verification, not more prose. An audit stopped by the time limit before it had written anything reportable is the one case that produces no report, and it says so.

06 Run

One click. RIGOR audits, verifies its own findings and builds the report.