Reconstruction evidence and limits
What was examined, how it was reviewed, what follows and what remains unproven.
Public scope note
Local reconstruction foundations - public scope note 1
- Evidence identifier
- reconstruction-scope-01
- Observed
- 2026-09-06
- Publication
- Candidate - not published
- Edition
- foundations-2026-09-06-preview-03
Editorial review of private implementation and internal acceptance material; not an external security audit.
- 1
What was examined
A bounded reconstruction implementation, separate oracle and engine paths, and their internal comparison and acceptance records.
- 2
How it was reviewed
Review of implementation scope, continuity and state-commitment checks, differential outcomes and the limits recorded with acceptance. The website does not rerun these private checks.
- 3
What the review supports
The reviewed material supports the limited claim that local reconstruction and comparison foundations are implemented. Agreement, refusal and disagreement are distinct observations; none is a blanket production-readiness verdict.
Claim ledger
Every public claim in this section, with the status the evidence supports, its basis and its limit.
| Claim | Status | Basis | Limit |
|---|---|---|---|
| Bounded reconstruction of a preserved history is implemented locally | Proven today | Internal implementation-scope review | Applies to the examined historical baseline |
| Two independent paths compare without default authority | Proven today | Differential review of oracle and engine outcomes | Separation alone does not exclude shared assumptions |
| Continuity and state-commitment checks are part of the local scope | Proven today | Internal implementation-scope review | Not a general database, a live network or unrestricted replay |
| Shared refusal classes and the one genuine asymmetric refusal are classified end to end | Proven today | Differential tribunal: both paths raise the paired class at the same unit; the sequence frontier is the one asymmetric surface reachable between them | Between two correct implementations no other asymmetric surface can occur without mutating a path |
| Every other asymmetric surface is classified only by synthetic observation | Qualified | Comparator-only table, executed from a hostile corpus, proved disjoint from and total with the end-to-end table | Qualified by construction: not input-reachable, and never converted into a winner |
| The result can be reproduced publicly | Later gate | Not yet available | Source and reproduction package remain private |
| The scope has been independently audited | Not claimed | None | Internal, agent-assisted review only |
| A distributed network, finality or a production service exists | Not claimed | None | Outside this evidence entirely |
Proven today, or practised on the lineIn development / qualifiedDirection: planned or researchLater gateNot claimed
What the evidence does not establish
- No distributed network, consensus finality, production throughput or financial-service availability is established by this evidence.
- A recorded acceptance applies to its examined scope and historical baseline. It does not certify the latest development branch.
- Asymmetric surfaces beyond the sequence frontier are classified by synthetic observation only and stay qualified by construction. A difference must not be silently converted into a winner or a conclusive diagnosis.
- Internal reviews, including agent-assisted reviews, are not independent external audits.
- A digest identifies bytes; it does not by itself prove that a process ran correctly.
Public reproducibility
The full source and private reproduction package are not public. This card is an explanatory scope note, not a downloadable proof or a publicly reproducible verification. The linked illustration uses synthetic data only.
Revision conditions
Re-review when the implementation baseline, accepted scope, known limits or availability changes. A withdrawn claim must also be removed from search and derived presentations.