docs(plans): record the remediation v2 close-out

[pacphi]

Card: #316 (A61-13) on the v4.0.0 board, milestone 4.0.0-alpha.61, epic #294. It blocks #317 and #320. Source: supersession ledger exit item 7.

Summary

Remediation v2 closes with the exit release alpha.61. This records that close-out in the plans, with each statement traced to a source in the repository. Archiving the plans is #317.

• The decision log has v2's summary entry (docs/plans/2026-09-26-issues-237-238-239-verification-and-decisions.md), the one the plan's V7 point 3 and the ledger's "Log close-out" row call for. It records what v2 delivered (V1–V7 reaching main through feat: integrate remediation v2 for human review #285 as 0511d575), where the open items went (each is a ledger exit item with a card on the board), and what the entry does not claim.
• DoD-3 is recorded, and cites the read-back run. The entry takes the first real dispatch from M1L:51-60: scheduled watch run 36729784908, three fired records at 14:31:47Z for AQE #735, #528 and #532, the push of commit 4704a58, and the GitHub notice read back (bot author, created 14:32:50Z). It states the limits as M1L does. This establishes recorded session creation, ledger publication and notice posting, not subscriber delivery or completed routine work. No open upstream-prefixed pull request was present at that read, so the outcomes are unverified.
• The v2 and EX Status sections are current. Both adopt the refreshes from 456ff58 exactly as written (applied with git apply, and they applied cleanly). Each also gains a 2026-10-02 close-out paragraph: v2 says it closes with alpha.61, and EX says Phase E is carried by chore: run or waive the maintainer's machine jobs #318 and chore: publish 4.0.0-alpha.61 #320, with the release checked through npx.
• Attended time is recorded as "not instrumented." v2's Status already said so. The decision-log entry now states it formally and claims no total.
• The global-match criterion is dropped from v2 §6 item 5, with a dated note pointing at the ledger's D-2 row. v2 §6 item 6 (DoD-3) gets a note pointing at the entry.

Things to know

• The "DoD-3" label comes from the ledger and the scope matrix, where it means v2 §6 item 6, not item 3. The entry says so.
• The entry's last paragraph is deliberate. It says the cards for fix(ci): Windows test legs take 8–15 min since #250 — aqe-store-merge.test.mjs is the critical path #262, chore: run or waive the maintainer's machine jobs #318 and chore: publish 4.0.0-alpha.61 #320 are open and that it claims none of the Windows timing evidence, the machine jobs or the release. I don't know whether the machine jobs have been done, so it only claims the card state.
• I did not edit the decision log's own Status section. It already says it archives alongside the program plans, which docs: archive the remediation and completion plans #317 does.
• The older text in v2 and EX is unchanged; the new Status paragraphs say it stays as written.

Tests

• pnpm run lint:md: 0 issues across 253 files. The first run caught a wrapped line that began with #735, which I reflowed.
• The docs-layout and GA-surface-guard suites pass, and every relative link in the new text points at a file that exists. I couldn't run lychee locally, so the links (internal) check in CI is the real one.
• Full pnpm test: 6,492 passed, 2 failed, 17 skipped. The two failures (process-tree, setup-memory-probe) fail identically on unmodified main in a container that runs as root, and pass in CI.

Closes #316

🤖 Generated with Claude Code

https://claude.ai/code/session_01WAbqdRJ4kk8UsjrKxYeUCd

---

Generated by Claude Code