Claims and
corrections.

Scope and non-goals

• CE-DEF-01 Replace project management or prescribe sprint cadence.
• CE-DEF-01 Eliminate human discussion or judgment.
• CE-DEF-01 Guarantee correctness or certify compliance.
• CE-DEF-01 Require autonomous execution.
• CE-DEF-01 Require a particular tool, vendor or file format.

Public claims register

CE-STATUS-01 Claim identifiers link important statements, diagrams, examples and metric definitions to this versioned register. Evidence quality E0–E3/X is separate from assurance class A0–A3. No entry asserts a local measured E2/E3 result.

Individual claim records

CE-DEF-01

design principle · E0 · current. The proposal complements adaptive delivery and links intent, authority, evidence, admission and outcomes.

Accountable duty: Method author. Last reviewed: 2026-10-20. A proposed synthesis, not a new invention or empirically established improvement.

CE-HYP-01

hypothesis · E0 · current. Shared state may reduce translation and rework enough to offset total upkeep.

Accountable duty: Method author. Last reviewed: 2026-10-20. No completed controlled or before/after pilot dataset is supplied.

CE-CTRL-01

design principle · E0 · current. The proposed profile forbids generator control of protected policy, signing and admission.

Accountable duty: Verification Owner. Last reviewed: 2026-10-20. A requirement; actual IAM/runner enforcement and graph completeness require independent implementation.

CE-LIFE-01

illustration · E0 · current. The local kernel checks declared lifecycle guards, expiry, expected revision and signed release bindings.

Accountable duty: Tooling owner. Last reviewed: 2026-10-20. Synthetic/local behavior only; no authenticated service, distributed store or deployment actuator.

CE-TRACE-01

design principle · E0 · current. Immutable event identities and explicit graph relations support traceable role-specific projections.

Accountable duty: Tooling owner. Last reviewed: 2026-10-20. The local hash chain is tamper-detecting given a trusted checkpoint; no durability, authentication or complete system discovery guarantee.

CE-OBS-01

observation · E1 · current. The author qualitatively observed less dispatch in a predecessor AWS setting.

Accountable duty: Method author. Last reviewed: 2026-10-20. Uncontrolled account; no measured baseline, denominator, effect size or model validation.

CE-X-01

established practice · X · current. External AI/delivery research motivates investigation and has context-specific causal limits.

Accountable duty: Method author. Last reviewed: 2026-10-20. External research does not evaluate the complete operating model.

CE-X-02

observation · X · current. GitClear observational churn analysis is not causal evidence for this operating model.

Accountable duty: Method author. Last reviewed: 2026-10-20. Projection and observed periods remain distinguished; code churn is not equivalent to customer harm.

CE-ILL-01

illustration · E0 · current. The token-expiry and routing walkthroughs illustrate requirements and checks.

Accountable duty: Method author. Last reviewed: 2026-10-20. Synthetic. The unsupported percentage-improvement narrative was removed in this revision.

CE-PILOT-01

design principle · E0 · current. Pilot windows, budgets, tier thresholds and review clocks are proposed planning parameters.

Accountable duty: Pilot lead to assign. Last reviewed: 2026-10-20. Preregister and adapt to risk; no universal affordability, adoption or performance threshold is established.

CE-METRIC-01

design principle · E0 · current. Metric cards define collection, formula, owners, counter-metrics and limits.

Accountable duty: Metrics owner to assign. Last reviewed: 2026-10-20. Definitions are not observations; historical coverage and comparable estimates remain unavailable.

CE-ASSURE-01

design principle · E0 · current. Assurance classes A0–A3 describe control diversity separately from publication evidence E0–E3/X.

Accountable duty: Verification Owner. Last reviewed: 2026-10-20. A declared classification does not prove statistical independence or installed authority separation.

CE-GOV-01

design principle · E0 · current. Named human responsibilities, temporary exceptions and protected escalation are required.

Accountable duty: Method author. Last reviewed: 2026-10-20. Pilot appointments and observed operational enforcement remain pending.

CE-PRACTICE-01

established practice · X · current. Adaptive delivery, ADRs, policy controls, provenance and independent certification have precedents.

Accountable duty: Method author. Last reviewed: 2026-10-20. The proposed linkage and its net benefit remain unproven; no superiority or first-invention claim.

CE-STATUS-01

observation · E0 · current. Operational completeness and independent adopter reproducibility remain unestablished.

Accountable duty: Method author. Last reviewed: 2026-10-20. No independent implementation study or E2+ pilot result is supplied; local tests establish only tested software behavior.

CE-PROJ-01

design principle · E0 · current. Purposeful audience projections need explicit owners, exact source lineage and drift checks; the local reference inspects supplied links, not semantic correctness.

Accountable duty: Method author. Last reviewed: 2026-10-20. Proposed mechanism and constructed local checks; no measured adoption cost, independent field implementation or pilot effect is established.

CE-ADOPT-01

design principle · E0 · current. Incremental learning can precede full record authoring while existing controls, baseline collection and release requirements remain mandatory.

Accountable duty: Method author. Last reviewed: 2026-10-20. Proposed mechanism and constructed local checks; no measured adoption cost, independent field implementation or pilot effect is established.

Corrections and material claim changes

Correction policy

CE-GOV-01 The method author owns public corrections. A report identifies the claim ID, exact version and supporting evidence. OpenKedge accepts correction reports through its existing public contact route; no response-time guarantee is asserted. Material corrections update the register, explain the reason and affected views, retain the old snapshot and link a versioned ADR. Named/identifiable case studies require participant authorization; confidentiality uses an approved anonymized report, never invented testimony. Remove unsupported performance figures pending review. Publish null, negative and inconclusive findings under the same method rules.

Search results and copied pages can retain older text. Current pages expose the canonical version and correction links; the project cannot purge third-party caches. Publication metadata distinguishes a staged local draft from an actually deployed release.

Readable revision history

Archived drafts

Archives preserve the bytes and stronger superseded wording for audit, not current endorsement. Earlier drafts without a recoverable snapshot are described in the review history rather than reconstructed and presented as authentic releases.

Limitations and open exits

CE-STATUS-01 Operational completeness requires an independent team to implement the specification without relying on its creator to interpret unclear semantics, observed protected runtime boundaries, a consented baseline/comparison and maintenance follow-up. No such independent implementation study or E2+ pilot is supplied. Named pilot appointments, representative verifier calibration and real cost/coverage measurements remain pending.