Eleven specialist agents review every preprint independently, then deliberate to produce a single Publication Fit tier (1–10) anchored to real journals.
— papers assessed across — indexed preprints from bioRxiv and medRxiv.
Machine-generated indicators. Assists but does not replace expert peer review.
Auto-refreshes every 30 seconds
Four layers. Eleven agents. One 10-tier Publication Fit score.
Paper-mill detection, statcheck p-value verification, GRIM tests, data availability, retraction cross-referencing — before any LLM runs.
Four integrity agents (methodologist, statistician, ethics, validity), five domain agents, a devil's advocate that stress-tests the consensus, and a positive-evidence reviewer that surfaces what the paper does well — all independent.
Evidence and significance labels from each agent map to a 2-axis grade via a rule-based lookup calibrated to eLife reviews, then derived to a 10-tier Publication Fit score anchored to real journals.
Borderline or low-agreement cases are arbitrated by Claude Opus with the full agent panel as context.
