THE CAPTURE REGISTRY IS CARD-RICH AND ORGANIC-POOR, AND THE PROFILE MAKES THAT THE FINDING. Of 588 observations across 428 addresses, 314 carry a structured card set — the sources a composition surface presented as its own — and the organic result set is noted in prose on 81 of those and structured on none. So the registry can say what entered the presented retrieval set, by source class, over time; it cannot yet separate archive-absent-organically from strong-organically-but-rejected-at-admission from admitted-but-recoded-in-composition. Those are three different phenomena, and the instrument that would distinguish them is a schema, not a theory.

WHAT THE CARDS SHOW ON GOOGLE AI OVERVIEW, 195 OBSERVATIONS, JUNE TO SEPTEMBER. Split by whether the entity is in the address: tied addresses admit the archive at a mean share of 0.74–0.86 in every month, with two zero-archive observations out of 101; untied addresses admit it at 0.26–0.53, with sixteen out of 94. The asymmetry the negative-ontology notebook (#1611) built on twelve captures holds across the whole carded record and does not move between months. The tied arm narrows in September: sources per observation fall from five to two or three while the answers lengthen (169 to 650 words at alexanarch symbolon, 57 to 386 at rebekah cranes sappho) and external sources per tied observation fall from 1.7 to 0.6 — longer answers from fewer sources, nearly all archive-owned. The untied arm's admitted substitutes change character: the external cards at concept addresses are "other web" in June and scholarly platforms, Wikipedia (28 in August from 1 in June), YouTube and code hosts by August. Read with the archive's own retrieval theory (#695 basin states, #172 the ambient attractor), that is an entity basin deepening and sealing while the competitor basins at the archive's concepts grow more institutional. […abridged for the catalogue; full description in this deposit's record]