Atlas of Kurpishev Artifacts

Abstract

This source-addressed Atlas presents the 36 Kurpishev Artifacts in the named Global RBD v2.2 registry, RUN-016. An artifact is treated as a registered Reper construction containing an identifier, assignments of the roles R, I, U, D, a relation graph, foundations, a proof stage and a route for further verification. All 36 passports are reproduced and connected to 16 underlying Reper nodes and 216 role-relation records. A separate analytical layer compares the catalogue with the static R3M-ENCODER-v1.0: seven distinct matrices produce two spectral passports, comprising 19 rank-one and 17 rank-two objects. An elementary result explains this distribution through the numbers of occupied relation families. Artifacts 1 and 18 demonstrate that identical matrices need not imply identical sufficient-foundation roles or proof stages. The Atlas connects the catalogue to the later complete-signature stage R4-K and source-admission stage R4-L without identifying the 36 artifacts with the Kant graph. The article records I. B. Kurpishev’s authorial architecture while separating cataloguing, mathematical analysis of the encoding, and empirical forecast validation that has not been performed.

Keywords: Kurpishev Artifacts; Reper; KLT-RBD; RIUD; Atlas; KASVD; proof passport; spectral collision; complete signature; sufficient foundation.

1. The artifact as a unit of scientific memory

Monograph 5.0 introduces event@state as the project’s elementary unit and uses a Reper to connect actual content, an internal invariant, a field of possibilities and a sufficient foundation. This construction is extended to sources, formulas, documents, programs and forecasts [2, §§1–3]. Volume XII develops it into the notion of an artifact: a recorded structure receives a source, domain, passport, status and next action [3, §§1–3, 12–13].

The Atlas retains the notation

$$A_n=(\mathrm{id}_n,G_n,\operatorname{Rep}_n,\mathrm{Dom}_n,D_n,S_n,B_n,N_n),\qquad \operatorname{Rep}_n=(R_n,I_n,U_n;D_n). \tag{1}$$

Here G is the graph, S the proof-state record, B the blocker list or field, and N the next verifiable step. This expression organizes the passport; it does not assert a universal numerical criterion of truth.

An artifact differs from an unrecorded idea through addressability and history. It can be located by its identifier, compared with an earlier version and checked for the foundations actually present. Its significance for the anthropology of cognition is that memory of an investigation becomes a reproducible object rather than only an account of a successful discovery. Preserved memory of a claim, however, is not a proof of that claim.

The article concerns the named registry 1–36. The broader Volume XII discusses mathematical, physical, chemical, biological and software artifacts. That thematic framework does not justify renaming these 36 particular RUN-016 records as 36 independent discoveries across all those disciplines. All 16 nodes directly used by the present catalogue have the preserved node-domain label KLT_RBD [1].

2. Source framework and preservation of identity

The primary source is KLT-RBD Global Reper Database v2.2, Artifact Naming and Dossier RUN-016. It contains the JSON/CSV name registry, 36 separate Russian dossiers, the Reper table, RUN-006 role relations and a SQLite database. The naming records are dated 2 June 2026. The number in “Artifact-n” equals the rank field of the RUN-015 queue: it is a ranked-candidate number, not an SVD rank or a discovery date [1, README; artifact_candidate_score_run_015].

The identifier lineage is

$$\text{RIUD/006}\to\text{PROOF/007}\to\text{FORMAL/009}\to\text{CAND/015}\to\text{NAME/016}. \tag{2}$$

This is a history of documentation and review. It is not, by itself, a time series of independent graph changes and must not be substituted for new observations in a prospective forecast.

All 286 checked entries of the original SHA-256 manifest matched during preparation of the Atlas, and ZIP CRC verification passed. The named JSON registry matched its SQLite table in every field; the role-relation table matched the CSV export. These checks establish the integrity of the archive snapshot used. They are not external scientific review or an independent EXT-RUN.

The four source families for the underlying nodes are kept separate: predictive specifications, five limit families, an interface catalogue and formal skeletons. The last source explicitly describes a formal-skeleton layer: a proof stub in a D field does not become a completed formal proof [1, sources_v2_2_preserved].

2.1. Reading numbers and statuses

The λ, δ_truth and CGI values are reproduced from RUN-016 and rounded to six decimal places in the article. They have not been recalculated from verbal R, I, U, D fields and are not presented as measurements of physical processes. The label δ_truth and scalar CGI are retained as historical registry fields. The later Volume V restricts λ to a strictly typed partial invariant and separates independent defect coordinates [4, pp. 1–2].

All 36 records carry proof_status = roadmap-only-not-proof and truth_layer_promotion = 0. Even the signature-ready decision for Artifact-1 denotes readiness for another controlled review, not a proved theorem. Artifact-2 has remaining_blockers = 0 but remains return-to-hold because an independent signature-witness is separately required. One numerical field cannot replace the other admission gates.

Ten cards without a formal_card_id contain 999 in remaining_blockers. This value is preserved literally and marked by an asterisk. It is not interpreted as an enumerated list of 999 independent obstructions. Likewise, d_coverage = 4 for every object is a recorded coverage characteristic, not four independently proved sufficient foundations.

3. Sixteen underlying nodes and the passport structure

Every card assigns the roles R, I, U, D to four distinct Reper nodes. A node’s own internal R, I, U, D fields must not be confused with that node’s role within an artifact: these are different descriptive levels. Compact locators N001…N131 are used for reading; they do not replace the original KRP-V20 identifiers, which are expanded in the machine-readable registry.

The first five nodes concern limits; the next six concern predictive records and reconstruction; three concern interface operations; and the last two concern formal skeletons. The last column gives short explanations of the source concept/concept_ru entries. Complete source wording, internal D fields, source_id and work_id are retained in the accompanying JSON, while the original bytes are preserved in the unchanged archive.

Each RIUD_TETRA contains six role pairs. “Tetrahedron” here identifies a four-vertex relational skeleton. It does not mean that the source specifies Euclidean coordinates, measured edge lengths or a geometric volume. A role record’s active_relation_signature may contain several directed relations; the six passport rows therefore cannot simply be called six unique edges of the entire global RBD.

Appendix A presents all 36 cards in the original naming order. RUN-006 object, RUN-007 proof-object and RUN-009 formal-card identifiers are abbreviated only in print; the expansion convention is stated before the catalogue.

4. From the Reper graph to a static SVD passport

The Atlas reproduces R3M-ENCODER-v1.0 [5]. Its rows represent RI, RU, IU, while its columns represent three relation families:

Matrix entries indicate whether at least one relation in a family occurs in a pair’s signature:

$$M_{p,f}=\mathbf1\{\operatorname{signature}(p)\text{ contains a relation in family }f\},\quad p\in\{RI,RU,IU\}. \tag{3}$$

The matrix uses the three pairs not involving D. The remaining three pairs, sufficient-foundation field, blockers and proof_status are not removed from the passport: they remain independent source and proof layers. Direction and multiple relations within the same family are not distinguished by this particular binary projection. That is an explicitly chosen compression, not a property of the source itself.

The R3-M matrix must not be mixed with the later R3-V demonstration matrix whose axes are R,I,U and D,P,G. These are different encoders. Every numerical SVD result in this article concerns the former contract only.

For M, the standard decomposition M = UΣVᵀ is used; LAPACK provides the classical algorithmic and terminological background [9]. The spectral passport is

$$\sigma_1\ge\sigma_2\ge\sigma_3\ge0,\qquad \rho=\|M\|_F,\qquad s_2=\sigma_2/\sigma_1,\quad s_3=\sigma_3/\sigma_1 \quad(\sigma_1>0). \tag{4}$$

Recalculation confirmed seven distinct matrices. The labels M1…M7 are introduced solely for navigation in this article, not as new artifact numbers.

Proposition 1. Why this Atlas has only two spectra

Suppose every row of a binary matrix contains exactly one 1. Let n_F, n_E, n_L count the rows assigned to each family. Then

$$M^TM=\operatorname{diag}(n_F,n_E,n_L),\qquad \{\sigma_i^2\}_{i=1}^3=\{n_F,n_E,n_L\}. \tag{5}$$

Proof. No row contains a 1 in two different columns, so every off-diagonal entry of MᵀM is zero. Each diagonal entry counts the 1s in its column. Singular values are the square roots of the eigenvalues of MᵀM. □

The row condition was checked for all 36 source objects. Family counts have only the forms (3,0,0) and permutations of (2,1,0). Therefore

$$19\times(\sqrt3,0,0),\qquad17\times(\sqrt2,1,0),\qquad \rho=\sqrt3\ \text{for every card}. \tag{6}$$

In the KASVD triangle these are (0,0) and (1/√2,0), explaining the distribution of 19 S1 and 17 S2g cards. It does not prove that the artifacts have only one or two semantic degrees of freedom: the calculation measures relation-family distinguishability within a fixed projection.

5. Two cards, one matrix: why foundations remain separate

Artifact-1 and Artifact-18 provide a particularly clear example. In both cards, R is forecast_fact_gap and I is nearest_rebuild_neighborhood. However, the former has U = LimitOK and D = PredReperRecord, whereas the latter has U = PredReperRecord and D = predictive_health. Their RUN-016 decisions are signature-ready and hold, and remaining_blockers is 0 and 6, respectively [1, dossiers 001 and 018].

Nevertheless, both yield

$$M_1=M_{18}=\begin{pmatrix}0&1&0\\1&0&0\\1&0&0\end{pmatrix}. \tag{7}$$

The three selected pairs enter the same families, while the three relations involving D differ. Matrix equality means equality of the chosen projection, not identity of the passport, proof or content. It does not authorize merging the database cards.

Artifacts 2 and 3 give another example: their matrices coincide, but their U roles refer to predictive_limits and predictive_forecast, respectively. A static spectrum need not distinguish a limit-catalogue request from a forecast request. The R4-J line separates encoder collisions from matrix-orbit matches for precisely this reason [6].

6. Connecting the Atlas to complete signatures R4-K/R4-L

The later mathematical line does not replace the source names. It refines the comparison of registered structures. Given an explicit finite complex with δ₁δ₀ = 0 and a fixed metric, it considers

$$H^1=\ker\delta_1/\text{im}\delta_0,\qquad h([z])=P_{\mathcal H}z. \tag{8}$$

The classical finite-dimensional Hodge framework is described by L.-H. Lim [8]. R4-K establishes complete comparison for a fixed finite group Γ of admissible isometries through a canonical orbit representative:

$$c_\Gamma([z])=\min_{g\in\Gamma}^{\mathrm{lex}}Q_g h([z]). \tag{9}$$

This formula does not require SVD to be a complete invariant of the original class. It does require justification of the complex, numerical defect z, metric and admissible group. The present article does not perform such a semantic cohomology calculation for all 36 artifacts. The static R3-M matrix is not presented as its substitute.

R4-L studied a different object, KANT-SNAP-015: 55 nodes, 59 separate edges, six components and ten cycle coordinates of the one-dimensional skeleton. It preserves an explicit witness of information loss when two edges with identical endpoints are merged. This is a methodological example for Atlas/Analyzer development, not Artifacts 37–91 or a renaming of the 1–36 catalogue [6, R4-L]. The same report leaves the semantic defect vector inadmissible: a missing value is not replaced by zero.

The common architectural route is

$$\text{source}\to\text{passport}\to\text{typed graph}\to\text{admissible class}\to\text{complete signature}. \tag{10}$$

KASVD accompanies it as a diagnostic spectral projection. The full graph and foundations must remain recoverable.

7. Role in the predictive method and the boundary of novelty

The Atlas strengthens the predictive Reper projective-harmonic method in three ways. First, artifact names become addressable objects with preserved histories. Second, it provides a verifiable comparison of graph forms and identifies the limits of compression. Third, it associates each stage with a specific next action rather than a promised outcome: obtain an independent witness, clarify a domain, rebuild a role map or repeat a review.

For Analyzer, this supports separating structurally similar fragments with different proof foundations. For ARBD, it requires retaining distinct versions and typed edges rather than merging them on spectral equality alone. For KLT-MATDES, the method can be transferred only after defining domain roles, units, admissibility and Evidence-D. The Atlas does not claim that these changes have already been integrated into production applications.

I. B. Kurpishev’s authorial contribution in this context is the architecture connecting artifact, Reper, source, D/Dom, graph, status and verification; its numbered machine-readable registry; and the integration of spectral and complete-orbit analysis into KLT-RBD. SVD, graph theory, matrix rank, Hodge theory and standard group actions are not claimed as authorial discoveries. External priority for every individual result is not established by cataloguing.

A numerical λ-passport does not replace proof; a populated D role does not establish sufficient foundation; signature-ready does not mean proved theorem; matching spectra do not establish artifact identity. These boundaries agree with PILOT-01 and Volume V [4; 7]. The semantic question remains decisive: what is transported between two records, and what justifies that transport?

Conclusion

A unified Atlas has been assembled for all 36 named artifacts in the RUN-016 snapshot. The registry is not reduced and its scientific statuses are not promoted. The source layer contains 36 dossiers, 16 underlying nodes and 216 role records. The analytical layer contains seven matrices, two spectra and explicit cases where identical matrices accompany different full passports. The Atlas is an intermediate publication stage after R4-L; the next substantive step remains R4-M, the source-level analysis of two parallel Kant relations and the foundations of their comparison.

Predictive-validation statuses: SCIENCE = ABSTAIN; EXT_RUN = NOT_RUN; truth_layer_promotion = 0. Mathematical calculations on the catalogue do not change these fields.

References and source locations

[1] Kurpishev I. B. KLT-RBD Global Reper Database v2.2: Artifact Naming and Dossier RUN-016. 2026-06-02. README_RU.md; exports/json/artifact_kurpishev_name_registry_run_016.json; dossiers/artifacts/ARTIFACT_001…036_KURPISHEV_RU.md; exports/csv/repers_v2_2_preserved.csv; tetra_role_edges_run_006.csv. Archive: sources/RUN016_ORIGINAL.zip.

[2] Kurpishev I. B. Monograph 5.0: Kurpishev Logic. 2026. monograph5_0_ru.pdf, pp. 3-4, sections 1-3. Original Russian title: Monografiya 5.0: Logika Kurpisheva.

[3] Kurpishev I. B. Volume XII: Kurpishev Artifacts and the Architecture of Mathematics. mbp179, 2026, pp. 3-7, sections 1-3 and 12-13. Preserved within v32_v196cc.pdf, PDF pp. 1109-1113.

[4] Kurpishev I. B. Basic Project KLT-RBD. Volume V: Computational Architecture. T5-AUDIT-MATH-v0.4, 2026-08-02, pp. 1-2 and chapters 4-9. tom5_v0_4_0_1(1).pdf.

[5] Kurpishev I. B. R3-M: Real Artifact Encoder / Admission Audit v1.0. 2026. R3M-ENCODER-v1.0; R3M_ARTIFACT_SVD_ATLAS_v1_0.json; r3m_build_static_atlas.py. Static matrix/spectrum record, not a temporal trial.

[6] Kurpishev I. B. R4-J: Quotient Spectral Geometry; R4-K: Complete Obstruction Signature; R4-L: Source-bound Signature Admission. Project checkpoints, 2026. R4L_RESULTS.json; KANT-SNAP-015 is a separate source graph.

[7] Kurpishev I. B. Reper-Projective Architecture of Formula Chains: PILOT-01. 2026-05-03, pp. 3-5, Definitions 3.1-3.3 and Theorem 4.1. PILOT01_PREPRINT_RU_FINAL.pdf.

[8] Lim, Lek-Heng. Hodge Laplacians on graphs. arXiv:1507.05379. https://arxiv.org/abs/1507.05379. Classical background for finite cochain complexes and Hodge representatives.

[9] LAPACK Users’ Guide. Singular Value Decomposition (SVD). Netlib, section 2.2.4. https://www.netlib.org/lapack/lug/node32.html. Classical definition and computational background.

Appendix A. Complete catalogue of Artifacts 1–36

Every record below has object_type = RIUD_TETRA, proof_status = roadmap-only-not-proof, internal_edge_count = 6, d_coverage = 4 and truth_layer_promotion = 0. Values of λ, δ and CGI are RUN-016 snapshot fields rounded for print. B reproduces remaining_blockers; 999* is not treated as a verified list of 999 items. Abbreviated numbers expand as follows: object → RIUD-TETRA-RUN006-nnn; proof → PROOF-OBJ-RUN007-nnnn; formal → FORMAL-CARD-RUN009-nnn. A dash denotes a missing formal_card_id. Complete identifiers and source wording are preserved in ATLAS_REGISTRY.json and the original dossiers [1].

A.1. Artifacts 1-6

01. Kurpishev Artifact-1

R = N017 (Forecast–fact gap); I = N018 (Nearest reconstruction neighborhood); U = N131 (LimitOK — formal skeleton); D = N130 (PredReperRecord — formal skeleton).

Object 029; proof 0029; formal 006. λ = -0.992737; δ = 0.007263; CGI = 0.140949; B = 0. M1 / S2g.

signature-ready. Next action: Use as reference; prepare the article and pass publication gates without changing the truth layer.

02. Kurpishev Artifact-2

R = N018 (Nearest reconstruction neighborhood); I = N019 (Search in semantic fog); U = N123 (Limit-catalogue interface); D = N130 (PredReperRecord — formal skeleton).

Object 030; proof 0030; formal 002. λ = -0.977250; δ = 0.022750; CGI = 0.134613; B = 0. M2 / S1.

return-to-hold. Next action: Obtain an independent signature-witness and repeat signature review.

03. Kurpishev Artifact-3

R = N018 (Nearest reconstruction neighborhood); I = N019 (Search in semantic fog); U = N124 (Predictive-Reper interface); D = N130 (PredReperRecord — formal skeleton).

Object 031; proof 0031; formal 001. λ = -1.035478; δ = 0.035478; CGI = 0.134599; B = 1. M2 / S1.

return-to-hold. Next action: Resolve the remaining blocker and repeat accept-review closure.

04. Kurpishev Artifact-4

R = N017 (Forecast–fact gap); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 011; proof 0011; formal 019. λ = -1.028786; δ = 0.028786; CGI = 0.160747; B = 6. M2 / S1.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

05. Kurpishev Artifact-5

R = N017 (Forecast–fact gap); I = N124 (Predictive-Reper interface); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 010; proof 0010; formal 015. λ = -1.035035; δ = 0.035035; CGI = 0.158355; B = 6. M2 / S1.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

06. Kurpishev Artifact-6

R = N015 (Event–state–limit packet); I = N003 (Spatial limit); U = N016 (Admissible future Reper); D = N002 (Ecological limit).

Object 028; proof 0028; formal 007. λ = -0.940909; δ = 0.059091; CGI = 0.141894; B = 4. M3 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

A.2. Artifacts 7-12

07. Kurpishev Artifact-7

R = N015 (Event–state–limit packet); I = N002 (Ecological limit); U = N016 (Admissible future Reper); D = N001 (Political limit).

Object 034; proof 0034; formal 003. λ = -0.929507; δ = 0.070493; CGI = 0.139731; B = 4. M3 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

08. Kurpishev Artifact-8

R = N015 (Event–state–limit packet); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 001; proof 0001; formal 023. λ = -1.075021; δ = 0.075021; CGI = 0.163651; B = 6. M2 / S1.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

09. Kurpishev Artifact-9

R = N017 (Forecast–fact gap); I = N018 (Nearest reconstruction neighborhood); U = N123 (Limit-catalogue interface); D = N130 (PredReperRecord — formal skeleton).

Object 020; proof 0020; formal 005. λ = -1.070872; δ = 0.070872; CGI = 0.140114; B = 5. M2 / S1.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

10. Kurpishev Artifact-10

R = N015 (Event–state–limit packet); I = N003 (Spatial limit); U = N016 (Admissible future Reper); D = N004 (Temporal limit).

Object 027; proof 0027; formal 010. λ = -1.077951; δ = 0.077951; CGI = 0.155086; B = 5. M3 / S2g.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

11. Kurpishev Artifact-11

R = N015 (Event–state–limit packet); I = N123 (Limit-catalogue interface); U = N124 (Predictive-Reper interface); D = N130 (PredReperRecord — formal skeleton).

Object 023; proof 0023; formal 027. λ = -1.115256; δ = 0.115256; CGI = 0.170376; B = 6. M4 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

12. Kurpishev Artifact-12

R = N015 (Event–state–limit packet); I = N130 (PredReperRecord — formal skeleton); U = N016 (Admissible future Reper); D = N131 (LimitOK — formal skeleton).

Object 007; proof 0007; formal 022. λ = -1.124753; δ = 0.124753; CGI = 0.161702; B = 6. M5 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

A.3. Artifacts 13-18

13. Kurpishev Artifact-13

R = N015 (Event–state–limit packet); I = N003 (Spatial limit); U = N014 (Predictive Reper record); D = N004 (Temporal limit).

Object 033; proof 0033; formal 018. λ = -1.136651; δ = 0.136651; CGI = 0.159899; B = 6. M3 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

14. Kurpishev Artifact-14

R = N015 (Event–state–limit packet); I = N004 (Temporal limit); U = N016 (Admissible future Reper); D = N005 (Ontological limit).

Object 035; proof 0035; formal 009. λ = -1.142339; δ = 0.142339; CGI = 0.150880; B = 6. M3 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

15. Kurpishev Artifact-15

R = N017 (Forecast–fact gap); I = N018 (Nearest reconstruction neighborhood); U = N124 (Predictive-Reper interface); D = N130 (PredReperRecord — formal skeleton).

Object 019; proof 0019; formal 004. λ = -1.162476; δ = 0.162476; CGI = 0.140025; B = 6. M2 / S1.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

16. Kurpishev Artifact-16

R = N015 (Event–state–limit packet); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N122 (Predictive-layer availability).

Object 012; proof 0012; formal —. λ = -1.117153; δ = 0.117153; CGI = 0.251555; B = 999*. M2 / S1.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

17. Kurpishev Artifact-17

R = N015 (Event–state–limit packet); I = N130 (PredReperRecord — formal skeleton); U = N014 (Predictive Reper record); D = N131 (LimitOK — formal skeleton).

Object 008; proof 0008; formal 026. λ = -1.176892; δ = 0.176892; CGI = 0.169986; B = 6. M5 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

18. Kurpishev Artifact-18

R = N017 (Forecast–fact gap); I = N018 (Nearest reconstruction neighborhood); U = N130 (PredReperRecord — formal skeleton); D = N122 (Predictive-layer availability).

Object 021; proof 0021; formal 017. λ = -1.198979; δ = 0.198979; CGI = 0.159793; B = 6. M1 / S2g.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

A.4. Artifacts 19-24

19. Kurpishev Artifact-19

R = N015 (Event–state–limit packet); I = N122 (Predictive-layer availability); U = N014 (Predictive Reper record); D = N130 (PredReperRecord — formal skeleton).

Object 006; proof 0006; formal 030. λ = -1.198701; δ = 0.198701; CGI = 0.172854; B = 6. M2 / S1.

hold. Next action: Complete Dom, Evidence-D and proof blockers; return to board review.

20. Kurpishev Artifact-20

R = N015 (Event–state–limit packet); I = N123 (Limit-catalogue interface); U = N014 (Predictive Reper record); D = N122 (Predictive-layer availability).

Object 015; proof 0015; formal —. λ = -1.173009; δ = 0.173009; CGI = 0.261519; B = 999*. M2 / S1.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

21. Kurpishev Artifact-21

R = N015 (Event–state–limit packet); I = N003 (Spatial limit); U = N014 (Predictive Reper record); D = N002 (Ecological limit).

Object 036; proof 0036; formal 008. λ = -1.203088; δ = 0.203088; CGI = 0.146018; B = 9. M3 / S2g.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

22. Kurpishev Artifact-22

R = N017 (Forecast–fact gap); I = N122 (Predictive-layer availability); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 014; proof 0014; formal 021. λ = -1.214280; δ = 0.214280; CGI = 0.161548; B = 9. M2 / S1.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

23. Kurpishev Artifact-23

R = N015 (Event–state–limit packet); I = N122 (Predictive-layer availability); U = N123 (Limit-catalogue interface); D = N130 (PredReperRecord — formal skeleton).

Object 024; proof 0024; formal —. λ = -1.224723; δ = 0.224723; CGI = 0.258421; B = 999*. M4 / S2g.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

24. Kurpishev Artifact-24

R = N015 (Event–state–limit packet); I = N122 (Predictive-layer availability); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 005; proof 0005; formal 024. λ = -1.307519; δ = 0.307519; CGI = 0.164424; B = 10. M2 / S1.

reject. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

A.5. Artifacts 25-30

25. Kurpishev Artifact-25

R = N017 (Forecast–fact gap); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N122 (Predictive-layer availability).

Object 018; proof 0018; formal 028. λ = -1.346845; δ = 0.346845; CGI = 0.171850; B = 10. M2 / S1.

reject. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

26. Kurpishev Artifact-26

R = N015 (Event–state–limit packet); I = N124 (Predictive-Reper interface); U = N014 (Predictive Reper record); D = N130 (PredReperRecord — formal skeleton).

Object 004; proof 0004; formal 025. λ = -1.241361; δ = 0.241361; CGI = 0.169605; B = 9. M2 / S1.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

27. Kurpishev Artifact-27

R = N017 (Forecast–fact gap); I = N130 (PredReperRecord — formal skeleton); U = N016 (Admissible future Reper); D = N131 (LimitOK — formal skeleton).

Object 016; proof 0016; formal 016. λ = -1.258074; δ = 0.258074; CGI = 0.158730; B = 9. M5 / S2g.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

28. Kurpishev Artifact-28

R = N015 (Event–state–limit packet); I = N124 (Predictive-Reper interface); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 002; proof 0002; formal 020. λ = -1.280679; δ = 0.280679; CGI = 0.161340; B = 9. M2 / S1.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

29. Kurpishev Artifact-29

R = N015 (Event–state–limit packet); I = N123 (Limit-catalogue interface); U = N014 (Predictive Reper record); D = N130 (PredReperRecord — formal skeleton).

Object 003; proof 0003; formal 029. λ = -1.262708; δ = 0.262708; CGI = 0.172038; B = 9. M2 / S1.

rebuild. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

30. Kurpishev Artifact-30

R = N017 (Forecast–fact gap); I = N124 (Predictive-Reper interface); U = N016 (Admissible future Reper); D = N123 (Limit-catalogue interface).

Object 017; proof 0017; formal —. λ = -1.389975; δ = 0.389975; CGI = 0.255033; B = 999*. M2 / S1.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

A.6. Artifacts 31-36

31. Kurpishev Artifact-31

R = N015 (Event–state–limit packet); I = N124 (Predictive-Reper interface); U = N016 (Admissible future Reper); D = N123 (Limit-catalogue interface).

Object 009; proof 0009; formal —. λ = -1.334002; δ = 0.334002; CGI = 0.257608; B = 999*. M2 / S1.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

32. Kurpishev Artifact-32

R = N015 (Event–state–limit packet); I = N124 (Predictive-Reper interface); U = N014 (Predictive Reper record); D = N123 (Limit-catalogue interface).

Object 013; proof 0013; formal —. λ = -1.474022; δ = 0.474022; CGI = 0.267903; B = 999*. M2 / S1.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

33. Kurpishev Artifact-33

R = N124 (Predictive-Reper interface); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N130 (PredReperRecord — formal skeleton).

Object 022; proof 0022; formal —. λ = -1.397060; δ = 0.397060; CGI = 0.273415; B = 999*. M6 / S2g.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

34. Kurpishev Artifact-34

R = N130 (PredReperRecord — formal skeleton); I = N123 (Limit-catalogue interface); U = N016 (Admissible future Reper); D = N122 (Predictive-layer availability).

Object 025; proof 0025; formal —. λ = -1.409002; δ = 0.409002; CGI = 0.289066; B = 999*. M7 / S2g.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

35. Kurpishev Artifact-35

R = N014 (Predictive Reper record); I = N123 (Limit-catalogue interface); U = N124 (Predictive-Reper interface); D = N130 (PredReperRecord — formal skeleton).

Object 026; proof 0026; formal —. λ = -1.697617; δ = 0.697617; CGI = 0.353280; B = 999*. M4 / S2g.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.

36. Kurpishev Artifact-36

R = N014 (Predictive Reper record); I = N122 (Predictive-layer availability); U = N123 (Limit-catalogue interface); D = N130 (PredReperRecord — formal skeleton).

Object 032; proof 0032; formal —. λ = -2.053941; δ = 1.053941; CGI = 0.366055; B = 999*. M4 / S2g.

cgi-hold. Next action: Rebuild role map / proof closure; repeat the λ/CGI audit.