AWatch-rus: PR review evidence

Дата: 2026-06-23

pr_review_evidence_status: "pending_until_first_reviewed_pr_is_merged"

GitHub issue: https://github.com/igor04091968/AWatch-rus/issues/48

Этот документ фиксирует, что будет считаться evidence для PR-based review workflow. Он не утверждает, что external peer review уже выполнен.

Evidence Criteria

Первый evidence-backed reviewed PR должен содержать:

PR URL;
linked issue URL;
completed pull request template;
passed checks;
reviewer approval;
merge commit;
no bypass, or documented bypass with reason and follow-up checks.

Evidence Record

PR URL: pending
Linked issue URL: pending
Reviewer: pending
Reviewer type: pending
Approval URL or screenshot filename: pending
Passed checks: pending
Merge commit: pending
Bypass used: pending
Date: YYYY-MM-DD
Maintainer note: pending

Reviewer Interpretation

Review by the same maintainer improves change discipline but does not prove external peer review. External peer review must not be marked completed unless a reviewed public PR includes a reviewer who is not the submitting maintainer and the review is visible.

Current Status

PR workflow documentation: ready.
PR template: ready.
CODEOWNERS routing: ready.
First reviewed PR evidence: pending.
External peer review completed: not claimed.

Not Registry Release Evidence

PR review evidence is governance/process evidence for public development visibility. It is not registry release evidence and does not replace artifacts, checksums, logs or release evidence from the Russian build-runner.

Russian Contour Note

Primary registry-readiness contour remains Russian Gitea plus the planned Russian build-runner. GitHub remains public mirror validation only.

Guardrails

Do not publish secrets, private URLs, private account data or customer identifiers.
Do not claim completed external peer review until the evidence record is filled from a real reviewed PR.
Do not claim branch protection verification from PR evidence alone.
Do not claim completed registry submission.
Do not claim certification.
Do not claim SIEM/DLP replacement.
Do not claim ML/LLM-based detection.
Do not claim automatic remediation.

2.3 KiB Raw Blame History