docs(adr): ADR-0130 status is Accepted, in ADR-0125's spelling, dated by its landing merge - #15704

Part of #15453

Ruling A (maintainer, 2026-09-05, decision batch #40, verbatim 「同意」, recorded on the card by the director seat) executed on one record: docs/adr/0130-release-artifact-as-co-ownership-boundary.md. The other two records the ruling names are not in this PR — see Records not touched here.

The one field that changes

Before — the **Status**: field, soft-wrapped across source lines 3-5:

**Status**: Proposed (2026-09-01) — awaiting the maintainer's hand-merge, which is itself the
acceptance act for a governed surface (Prime Directive #14). ⛔ Nothing below is settled until
this record merges; the implementation cards are cut **from** the merged ADR, never ahead of it.

After — one line, in ADR-0125's spelling (its own line 3 was read as the template, not recalled):

**Status**: Accepted (2026-09-01) — accepted by the merge that landed it on `main` ([#14151](https://github.com/objectstack-ai/objectstack/pull/14151), commit `682d03ba7`), which is itself the acceptance act for a governed surface (Prime Directive #14).

1 insertion, 3 deletions in one file. The ruling names "line 3, its Status line"; in this record that field is soft-wrapped over source lines 3-5, so the whole field is replaced and nothing else. The trailing narrative goes with it because the ruling's rule is that the status line carries the state and the act and nothing else — and "Nothing below is settled until this record merges" is both narrative and false once the record is Accepted. Every other line of the file, including the 2026-09-02 addendum block that follows, is byte-identical.

The landing reading — the ORIGINAL landing, not the anchor migration

The ruling quoted 09cc6be43 for this record. That commit is the 2026-09-03 anchor migration; the card itself says the original landing is earlier and was not chased. It was chased here, in two independent channels that agree.

git — the clone is shallow, so it was deepened first. That mattered: on the shallow tree git log --follow answered with an unrelated 2026-09-04 commit. After git fetch --deepen:

$ git log --follow --diff-filter=A --format='%h %cd %s' --date=iso-strict -- docs/adr/0130-release-artifact-as-co-ownership-boundary.md
682d03ba7 2026-09-01T09:03:27+00:00 docs(adr): ADR-0130 — the release artifact is the co-ownership boundary (one artifact, N packages) (#14151)

REST — GET /repos/objectstack-ai/objectstack/commits?path=docs/adr/0130-release-artifact-as-co-ownership-boundary.md returns 3 commits for the path (09cc6be43 2026-09-03, 184f6b0ca 2026-09-03, 682d03ba7 2026-09-01); the oldest is 682d03ba7, and GET /commits/682d03ba7 reports files[0].status: "added" for exactly this path, with the message naming PR #14151 (the repo squash-merges).

Precondition set by the ruling — the landing PR's merged_by must be a human approver account. GET /repos/objectstack-ai/objectstack/pulls/14151, response fields quoted:

{
  "number": 14151,
  "merged": true,
  "merged_at": "2026-09-01T09:21:39Z",
  "merge_commit_sha": "682d03ba7e8325f8facd59bd71ece9fcb7ad6d3b",
  "base": { "ref": "main" },
  "user": { "login": "os-sam" },
  "merged_by": { "login": "os-zhuang", "type": "User", "id": 277994282 }
}

merged_by.login is os-zhuang — one of the human approver accounts the ruling names, not a seat account and not an app. So "the merge was the acceptance act" holds for this record, and the date is its landing day, 2026-09-01 (the commit date and merged_at fall on the same day).

Records not touched here

• ADR-0128 — its landing PR was merged by a seat account, not a human approver, which is exactly the case the ruling says it does not cover ("stop and report that record"). Reported on the card; ⛔ not edited.
• ADR-0131 — the precondition holds for it, but the file is in flight on another lane's draft PR, so it is a hard same-file serial and is a second flight after that PR lands or is closed.
• ADR-0126 — its **Original status**: history line is a grep near-miss, ⛔ deliberately untouched.
• No changeset (nothing publishes from this diff, skip-changeset); content/docs/releases/** untouched.

Derived surfaces — none needs regeneration

Searched for anything that pins this record's Status or the word Proposed for it: docs/adr/README* and docs/adr/index* do not exist (the only non-numbered file under docs/adr/ is PRIORITIZATION.md); no hit for 0130 under content/docs/** outside content/docs/releases/**; scripts/check-adr-*.mjs carry no status ledger. The only repo-wide reference to the file by path is examples/app-multi-package/README.md:3, a plain relative link with no line anchor, and check-adr-symbol-anchors reports 0 line anchors survive corpus-wide, so the 2-line shortening moves no anchor.

Gates

Family derived from the real change set, not from a hand-written list: node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack at commit 4e3ea9826 (1 path vs merge base 95d5cbb31) ⇒ 17 commands, all run, then reconciled with --ran:

✓ dispatch-gates --ran: 17 derived famil(ies) accounted for — 17 run, 0 NOT-MEASURED.

Two readings that are NOT MEASURED rather than green, stated so they are not misread:

• The four gates above that first answered exit 3 PREREQUISITE NOT MET (no node_modules, then @objectstack/formula / @objectstack/lint not built) were re-run after pnpm install and the two package builds; the table reports the re-runs.
• pnpm lint does not apply to this diff, and this is a measurement, not a skip: eslint --no-inline-config on the edited file, --format json, returns 1 result with errorCount: 0 and the single warning "File ignored because no matching configuration was supplied.", and eslint.config.mjs contains zero .md/markdown entries. The diff touches exactly one file, and it is one eslint is not configured to read, so no untouched file's verdict can move.

Governed-surface handling

docs/adr/** ⇒ draft PR, os-zhuang + hotlong requested, human merge. ⛔ Not flipped ready by the seat, ⛔ not enqueued, ⛔ no auto-merge, ⛔ no self-approval. Part of #15453 rather than a closing keyword, because two records on that card are still outstanding.

🤖 Generated with Claude Code

https://claude.ai/code/session_019RfFHiRCSs3JXLK4cwcfox

Generated by Claude Code