QAtlas — Verified Exact-Solution Atlas

Provisional v2 view — RES not wired

Generated by docs/atlas/generate.jl — a pure VIEW over the *_registry.jl claims + the static test/INVENTORY.jsonl AST scan. No test is executed and no src is run; test/INVENTORY.jsonl is regenerated in-place (idempotently) from that static scan; fetch/@register untouched. Assurance labels are PROVISIONAL: residuals / confidence are not shown yet (RES not wired). Badges reflect the committed test AST, not the latest CI run — a hub can read green while its @test is red between regenerations. @sweep = a graceful regime-resolution gap, not card omission.

Assurance taxonomy (R1, ratified — Discussion #379)

Five honest levels, highest achieved tier wins:

🟣 universality-corroborated — agreement checked against the universality class (exponents / central charge). Strongest, regime-spanning evidence.
🟢 corroborated-at-p — an independent computation (finite-size ED extrapolation or a second closed form) reproduces the value at concrete parameter point(s) p.
🔵 coherent — an independent in-repo card exists and the value satisfies an internal invariant (sum rule / limiting case / delegation / retype, or an unrecognised/missing route), but no external value re-derives it.
⚪ cited-only — backed only by a literature citation, or the model is ED-infeasible so a citation is the ceiling. The honest frontier — neutral, not a penalty.
🟠 uncorroborated-but-feasible — src claims the hub, dense ED is feasible, yet no card checks it. The only actionable risk.

Denominator split: the corroboration rate is taken over ED-feasible claimed hubs only. ED-infeasible models (AKLT2D, ChernSimons3D, FibonacciAnyons, KagomeHeisenbergAFM, KitaevHoneycomb, PpIp2DSC, SYK, ToricCode, XCube) are excluded from the risk denominator — their ceiling is the published / DMRG value.

Coverage (all models)

	count
Hubs `src` claims (registry)	394
ED-feasible claimed (risk denominator)	367
ED-infeasible claimed (frontier, excluded)	27
🟣 universality-corroborated	0
🟢 corroborated-at-p	177
🔵 coherent	43
⚪ cited-only (frontier — neutral)	19
🟠 uncorroborated-but-feasible (actionable risk)	155
Inventory cards scanned (whole test/)	2882
Registry files parsed	86 / 86
Models	69

Externally-corroborated rate (🟣+🟢 over ED-feasible claimed): 48.2% · in-repo-verified rate (incl. 🔵 coherent): 59.9%

Faceted search → · by model · by quantity · by BC · by level · by mechanism · by regime · by universality. Full-text search is the top bar (Documenter built-in).

Doc-health audit

Actionable gap surface — see Audit for the itemised list.

Section	Count
1. Models without CONVENTION header	4
2. Quantities without extracted Definition	64
3. Orphan calc notes (matched to no model)	0
4. Models registered but with 0 hubs	0
5. INVENTORY card hubs with no `@register` claim	10

model	claimed	🟢	🔵	⚪	🟠	ED
`AKLT1D`	18	15	0	1	2	feasible
`AKLT2D`	1	1	0	0	0	infeasible
`BCFT`	1	1	0	0	0	feasible
`Bound`	8	0	0	0	8	feasible
`ChernSimons3D`	2	1	0	1	0	infeasible
`Cluster1D`	2	2	0	0	0	feasible
`Compass1D`	1	1	0	0	0	feasible
`ConformalBootstrap`	1	0	0	1	0	feasible
`CurieWeissIsing`	10	6	0	0	4	feasible
`DMIHeisenberg1D`	1	1	0	0	0	feasible
`DimerLattice`	4	1	0	0	3	feasible
`ExtendedHubbard1D`	1	0	1	0	0	feasible
`FibonacciAnyons`	1	1	0	0	0	infeasible
`GrossNeveu`	2	2	0	0	0	feasible
`HaldaneShastry`	7	0	0	0	7	feasible
`Heisenberg1D`	26	12	4	0	10	feasible
`HeisenbergXYZ`	5	1	1	0	3	feasible
`Honeycomb`	2	0	0	0	2	feasible
`Hubbard1D`	5	2	1	0	2	feasible
`IsingChain1D`	9	7	0	0	2	feasible
`IsingSquare`	13	7	2	0	4	feasible
`IsingTriangular`	10	2	0	0	8	feasible
`J1J2Heisenberg1D`	1	1	0	0	0	feasible
`Kagome`	2	0	0	0	2	feasible
`KagomeHeisenbergAFM`	3	0	0	3	0	infeasible
`Kitaev1D`	9	5	0	0	4	feasible
`KitaevHeisenberg`	1	1	0	0	0	feasible
`KitaevHoneycomb`	10	1	5	4	0	infeasible
`Lieb`	2	0	0	0	2	feasible
`LiouvilleCFT`	2	2	0	0	0	feasible
`LogarithmicCFT`	1	1	0	0	0	feasible
`LongRangeIsing1D`	1	0	1	0	0	feasible
`LongRangeXY1D`	1	1	0	0	0	feasible
`MajumdarGhosh`	4	0	0	2	2	feasible
`MixedFieldIsing1D`	1	0	1	0	0	feasible
`PXP1D`	1	0	0	1	0	feasible
`PpIp2DSC`	2	2	0	0	0	infeasible
`RFIM`	1	1	0	0	0	feasible
`RandomBondIsing2D`	1	0	1	0	0	feasible
`RandomTFIM`	4	0	1	0	3	feasible
`RiceMele`	6	2	2	0	2	feasible
`S1AnisotropicD1D`	2	0	2	0	0	feasible
`S1Heisenberg1D`	19	10	3	1	5	feasible
`S1XXZ1D`	2	0	2	0	0	feasible
`SLEkappa`	2	2	0	0	0	feasible
`SSH`	9	5	0	0	4	feasible
`SYK`	1	1	0	0	0	infeasible
`SchwingerModel`	2	2	0	0	0	feasible
`ShastrySutherland`	1	0	0	0	1	feasible
`SherringtonKirkpatrick`	2	1	0	1	0	feasible
`SixVertex`	5	2	1	0	2	feasible
`SpinIce`	1	1	0	0	0	feasible
`TASEP`	2	1	0	0	1	feasible
`TFIM`	67	33	10	0	24	feasible
`TTbar`	1	1	0	0	0	feasible
`TightBinding1D`	19	4	0	0	15	feasible
`TightBindingV1D`	8	3	0	0	5	feasible
`TodaLattice`	1	1	0	0	0	feasible
`ToricCode`	6	3	0	3	0	infeasible
`Triangular`	2	0	0	0	2	feasible
`TricriticalIsing`	4	2	0	0	2	feasible
`TricriticalPotts3`	4	2	0	0	2	feasible
`Universality`	3	2	0	0	1	feasible
`XCube`	1	1	0	0	0	infeasible
`XXZ1D`	29	17	5	0	7	feasible
`XYh1D`	11	0	0	0	11	feasible
`YangLee`	3	1	0	1	1	feasible
`ZnClock`	2	1	0	0	1	feasible
`ZnParafermion`	2	1	0	0	1	feasible