docs(adr): point ADR-0041's References at where the trigger packages live

[os-warren]

Refs #15478. Carved out of PR #15993 — this is that PR's fourth file, on a branch of its own.

What

ADR-0041's ## References section names the two shipped trigger packages by source path. Both packages moved on 2026-06-12 and the pointer did not, so it names a directory that does not exist:

- - Shipped triggers: `packages/plugins/plugin-trigger-record-change`,
-   `packages/plugins/plugin-trigger-schedule`
+ - Shipped triggers: `packages/triggers/trigger-record-change`,
+   `packages/triggers/trigger-schedule`

Two lines, one file, +2/-2. The bytes are identical to what PR #15993 carried — the file's blob is 9016f38ff on both branches, so nothing was re-authored on the way over.

Why this is its own PR

docs/adr/** is a governed surface, so under Prime Directive #14 a diff that touches one is landed by the maintainer by hand and never through the merge queue. One path hit governs the whole PR — 「混合 diff 一条命中即整 PR 分叉」 — so this single line made PR #15993 ineligible as a whole, alongside two published-manifest corrections that are ordinary queue work.

That is measured, not predicted. On every added_to_merge_queue for PR #15993 (20:09Z, 21:49Z, 22:22Z, 01:57Z) the merge-group run's Governed Surface Queue Guard failed at the step "Governed surfaces may not enter the merge queue unreviewed" and github-merge-queue[bot] removed the PR. Four ejections. The PR-level check of the same name stayed green throughout, because that guard bites in the merge_group event — which is why 34 green PR checks never predicted any of them.

With this file carved out, PR #15993 is three paths with zero governed hits. Read from the register rather than recalled, and with its control fired in the opposite direction so the reading is a measurement and not an empty pass:

node scripts/pm/check-governed-merges.mjs --test  (the three remaining paths)
  EXIT=0   0 of 3 path(s) hit the register (5 surfaces, repo-agnostic)
           NOT governed — ordinary queue landing applies

node scripts/pm/check-governed-merges.mjs --test docs/adr/0041-flow-trigger-family.md
  EXIT=3   1 of 1 path(s) hit the register
           GOVERNED — docs/adr/** ×1

Landing

⛔ This PR waits for a maintainer's merge. It stays draft; no seat flips it ready, enqueues it, or arms auto-merge on it. That is where a docs/adr/** touch was always going to end up — the carve-out changes which PR waits, not whether one does.

Deliberately left as written

The same record's rename table and narrative at lines 46, 48 and 97-100. Those name the old identifiers as the decision's own history, and the table's left column is a package name, not a path — not a pointer by construction. The anchored replacement could not have reached them in any case, since they carry no packages/plugins/ prefix; the distinction is the point. A References pointer is live; a decision record's account of what it renamed is not.

No changeset

This publishes nothing from any package, so skip-changeset applies and is set on this PR.

Verification

Gate family derived mechanically on head c18ae4060 — node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --commands, which reads the change set from the merge base itself (1 path), never hand-built. Every exit code captured immediately after a single redirected command, never through a pipe.

All 17 derived commands EXIT=0: check-adr-links (and its self-test), check-adr-symbol-anchors (and its self-test), check-ci-filter-parity, check-closing-keyword-parity (and its self-test), check-comment-mask-corpus, check:doc-formula-expressions, check:adr-anchors, check:cross-package-test-inputs, check:doc-authoring, check:driver-memory-census, check:nul-bytes, check:pm-governed-merges, check:refd-timer-probe, check:watch-hint-literal.

⚠️ One of those needed its prerequisite met first and is not reported as a first-pass pass: pnpm --filter @objectstack/lint run check:doc-formula-expressions answered 3 = PREREQUISITE NOT MET (unbuilt @objectstack/formula and @objectstack/lint) — neither green nor red, nothing measured. After turbo run build --filter=@objectstack/formula --filter=@objectstack/lint (4/4 tasks) it re-ran EXIT=0.

⚠️ pnpm check:pm-governed-merges runs the checker's own --self-test (274 assertions), so it grades the checker and not this diff. The verdict on this diff is the --test pair quoted above.

⛔ Nothing here is a claim about CI. Working tree clean at c18ae4060.

🤖 Generated with Claude Code

https://claude.ai/code/session_01XpTx2tbq3pZRYAdoGt6E6Y

---

Generated by Claude Code