Corpus Engine
Literature triage and assay census
Triage the literature corpus
Find every paper an automated platform could actually reproduce.
The engine classifies each record as data-only, physical or mixed, extracts its assay vocabulary, counts what the whole corpus uses, and ranks candidates by how cheaply the reproduction would be.
Classify
data_only, physical or mixed, with a calibrated confidence and its reasons.
Census
Assay usage across the corpus, with Wilson intervals on every share.
Rank
The cheapest replications first, and an explicit path to everything else.
Sign in
Open the corpus workspace.