Historical 0.2.7a0 R0 release-readiness record

This page is a preserved pre-stable audit record. It is not the current public release status; use the active 0.2.7 verification summary for the current conclusion.

R0_STATUS = PASS_WITH_LIMITATIONS.

R0 starts from the clean F6 source commit 5323a0996f214a3203f06b5bb468843b57c25270. The release worktree initially contained one unrelated modification to docs/verification/project_hygiene_architecture_audit_0_2_1.md. It was classified as FOREIGN_CHANGE and preserved in a targeted stash with commit a9e64d5e03ebadfcd1292ad3c5b04460761c10aa; it is not part of the candidate.

The current candidate remains 0.2.7a0, with no tag, GitHub release or PyPI publication. The future tag/build source is the final local R0 closeout commit reported with this record. The machine-readable record is qualification/0_2_7/r0_release_readiness.json.

Gate result

Area	Result	Evidence
Source integrity	PASS	Clean release tree after foreign-change isolation; no numerical or baseline change
Qualification integrity	PASS_WITH_LIMITATIONS	F1-F6 closed; active accounting remains 96/100
Audit closure	PASS_WITH_LIMITATIONS	Three unchanged F4/F6 failures remain outside the bounded public matrix
Test readiness	PASS	138 passed, 2 skipped targeted closeout battery; static checks pass
Packaging	PASS_WITH_LIMITATIONS	Rebuilt wheel/sdist pass distribution and twine checks; clean-install evidence is inherited from F5
Documentation	PASS_WITH_LIMITATIONS	Candidate, boundaries and known limitations are explicit
Publication traceability	OWNER_ACTION_REQUIRED	Generic readiness script correctly remains NOT_READY before tag and manual history review

The public source/archive audits pass. The generic readiness script still reports git_history_audit and version_tag as failed checks. This is expected before publication: the history prefilter requires a manual owner review of a legitimate public controlled-proof path and a historical non-GitHub-no-reply author identity, while R0 intentionally does not create the tag. No secret or local credential was found in the flagged proof contents.

The 5M Silver evidence remains applicable to the recorded structured TET4 PETSc/MPI workload. The C3 10M evidence remains bounded and is not a universal capacity or scaling claim. WEDGE6 static remains EXPERIMENTAL; WEDGE6 modal retains only its separate bounded scope. CalculiX remains NOT_COMPARABLE for the bounded WEDGE6 contract. Optional environments, macOS/Python-version coverage, and the three experimental/stale nonlinear full-suite failures remain known limitations.

R0 does not change numerical source, active baselines, maturity, historical evidence or release version. It does not push, tag, publish, or create a GitHub release. The owner may decide whether to proceed after the explicit manual history and license review required by the publication tooling.

RECOMMENDATION = RELEASE_WITH_KNOWN_LIMITATIONS.