chore: restore post-v0.2.3 development identity

[YuzeJ21]

Summary

• advance the post-release source identity from published v0.2.3 to unreleased 0.2.4.dev0 while leaving the v0.2.3 tag, GitHub Release, assets, checksums, and commit-bound historical evidence intact;
• add a fail-closed repository contract that prevents a published final version from being reused by a different committed tree, tracked/index change, or package-relevant untracked input—including ignored files under force-included wheel roots—while preserving explicitly build-excluded local state such as .coverage 2;
• record post-v0.2.3 CLI lifecycle parity, strict saved-record validation, packaged Chromium coverage, Python 3.13, bounded keyboard/focus and zoom evidence, and verified-public provenance under Unreleased;
• reconcile README, roadmap, development environment, v0.2.3 status and platform matrices, market comparison, and superseded audit boundaries;
• remove implemented CLI lifecycle and packaged-browser work from future proposals while keeping real screen-reader, Windows desktop, Linux desktop, non-Chromium, and WCAG evidence unsupported.

Repository truth

• Base: 077f9351283b319b82854ad1df95eac7ce614e21
• Published v0.2.3 release commit: 448c42758ea139bf9203cbf1bb04b02b02ae412c
• Development identity: 0.2.4.dev0
• Post-release engineering recorded from merged PRs feat: add post-release CLI lifecycle and packaged browser proof #185, test: add accessibility and Python 3.13 evidence #187, and fix: enforce verified public repository provenance #188
• Stage 1 remains exactly 0/5 qualifying reviews, 0/3 practitioners, 0/3 public repositories, 0/3 independently observed under-ten-minute completions, and 0/2 reuse-intent signals.

TDD and verification

• Observed the release-tree contract fail before changing 0.2.3, then pass after advancing the version.
• Added and observed failing regressions for tracked dirty state, package-relevant untracked files, the real non-ignored .coverage 2 exclusion, and ignored files under Hatch force-included wheel roots before implementing each repair.
• Ruff: passed.
• Complete suite: 1,993 passed and 2 intentional skips.
• Combined coverage: 95.06%; the 95% gate passed.
• Repository contracts: 84 passed.
• Acceptance benchmark: 12 cases / 13 criteria; zero mismatches, zero must-have False Ready outcomes, zero false blockers, and zero unexecuted categories.
• Comparison benchmark: 2 cases; zero mismatches.
• Two exact-head wheel builds with SOURCE_DATE_EPOCH=1706745600 were byte-identical at SHA-256 66a8843307aefae85da21e7b678c513ebaacf00a8b3c580a0446a3075a6fa076.
• Artifact inventory: 101 wheel entries, 556 source-distribution entries, zero forbidden matches.
• Clean install: pip check passed; distribution, module, and new-review identities all reported 0.2.4.dev0; both CLI versions, both installed benchmarks, and exact loopback health passed.
• Installed-wheel Chromium regression: 3 passed with the existing loopback-only networking and console/page-error assertions.
• Independent review: all four actionable Important findings were repaired test-first; final re-review found no remaining Critical or Important findings.

Evidence boundary

This is ScopeProof engineering evidence only. ScopeProof remains an evidence assistant, not a correctness oracle, and does not execute target-repository code. Static candidates are not runtime verification. Reviewer/source-owner identity remains asserted rather than authenticated. The GitHub Action remains opt-in and informational. No release, tag, package publication, outreach, R-002 retuning, R-003 generation, beta activation, private-repository support, or Stage 1 credit is included.

Real screen-reader operation, Windows desktop, Linux desktop, non-Chromium browsers, WCAG conformance, Python 3.14 compatibility, and genuine external adoption remain unsupported or unverified.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: eb7df35183

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".