[objectstack-fleet]

Requested by Tesla · project thread

Fixes #27521

Clause-②: no
Metadata-protocol: none

维护者速读

• 改了什么：平台测试清单 studio-authoring.first-run-loop 第 4 步（steps[3]）里引用的发布提示，从已下线的「Published 1 item in this package (one atomic release)」改成控制台现在实际显示的文字（按数量的「Published N items in this package」，发布对象时附带权限集提示和「Open Access」按钮）；条目 revision 8 → 9，并追加一条 history。
• 为什么改：Setup and Studio: suggestions from the 18.3.0 RC browser walkthrough: Setup naming and machine values, invitation and user onboarding, permission-set authoring, dialog focus, Studio first run, validation and flow builders, and publish-panel wording #27521 第 28 行已把「atomic」从发布标题、提示和确认说明里去掉，但这份 QA 清单仍引用旧文字，测试员按清单核对会判错。这是该卡最后一个可在 ui 车道完成的项；第 9 行（取消邀请确认框的「Keep invitation」按钮）需要 spec 新键，已另立 Action confirm and parameter dialogs: a declared dismiss label (ActionSchema.cancelText), so "Cancel Invitation" no longer sits beside "Cancel" #28519 承接。
• 风险与回滚：只改一份 QA 文档 JSON 的一句步骤文字和修订记录，不改代码、契约或任何发布内容；回滚即 revert 本 PR。
• 先看：docs/qa/platform-checklist/areas/studio-authoring.json 第 31 行（steps[3]）和第 96 行（revision 9 的 history）。
• 两分钟验证：pnpm check:platform-checklist 退出 0；对照 ui/packages/i18n/src/namespaces/studio/en.ts 的 engine.studio.publishedAllCount* 两个键，文字一致。
• 需要你做的：无，席位落地。

What changes

• docs/qa/platform-checklist/areas/studio-authoring.json, item studio-authoring.first-run-loop, steps[3]: the publish toast is quoted as main draws it. It read "Published 1 item in this package (one atomic release)", wording PR fix(studio): plainer authoring in Studio — option labels first, named rules, labelled fields, honest counts #27830 retired for this card's row 28. It now reads "Published N items in this package" ("Published 1 item in this package" for one; keys engine.studio.publishedAllCount / publishedAllCountOne), with the unchanged "Published all drafts in this package" fallback. Because that step's publish puts an object live, the step also names the hint the toast adds since row 35 ("Members can use it once a permission set grants them access.") and its "Open Access" action (StudioDesignSurface.tsx).
• Revision 8 → 9 with a history entry. Wording only; no clause changed.
• Lines 429, 464 and 527 of the same file still say "atomic". They describe the server's publish-drafts semantics or quote the 2026-07 audit document, and both are still accurate, so they stay.

Why

This closes #27521. Every row of the card now has a disposition:

• Row 9 (admin half): the declared Cancel Invitation confirm still reads "Cancel" / "Cancel Invitation". ActionRunner.confirmOptions() puts the action label on the confirm button, and ActionSchema has no key for the dismiss label. Seat pick: the spec card Action confirm and parameter dialogs: a declared dismiss label (ActionSchema.cancelText), so "Cancel Invitation" no longer sits beside "Cancel" #28519 (ActionSchema.cancelText, then sys_invitation declares "Keep invitation") takes the row.
• The leftover named in the Studio rows' os-dev report (this docs quote) lands here.
• The other 38 rows were already disposed. The row ledger is below.

Row ledger (measured on main efb1620 from the card body, all comments, and the PR #27830 / #27849 row tables)

Rows	Disposition
1, 20, 21	landed in PR #27849
2	card #27853 (remainder #28069)
3	card #27854
4	info colour landed in PR #27849 · rest card #27855 (domain:server, open)
5	keyboard tooltip landed in PR #27849 · hover already on main · empty states will-not-do
6	card #27797
7, 8, 23	landed in PR #28224
9	Resend confirm and the workspace "Keep invitation" landed in PR #28224 · admin dismiss label goes to #28519
10	invite half landed in PR #28224 · Create User half card #27857
11	sign-in hint and joined state landed in PR #28224 · pre-fill will-not-do (ruling 6545261914)
12, 13, 22	ui halves landed in PR #27849 · rest cards #27858, #27859, #27869 (and #28107, domain:server, open)
14	cards #27861, #27862
15	hover reveal will-not-do · columns card #27863
16	name derivation already on main · cards #27864, #27865
17, 19	cards #27866, #27867 (and #28068)
18	already on main via #27808
24, 31, 36	cards #27833, #27836 (and #27895), #27837 (and #27904, domain:engine, open)
25, 26, 32, 33, 35, 38	landed in PR #27830
27	defect half card #27799 · suggestion half card #27834
28	wording landed in PR #27830 · naming card #27835 · the QA quote in this PR
29, 37	will-not-do (PR #27830 table)
30, 34	landed in part in PR #27830 · header truncation will-not-do · Publish reason already shipped
39	card #27838 (and #28234, domain:server, open)

Verification

Derived gates (head 5e94fc0)

node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack derived 16 families. Each ran with stdin from /dev/null, exit codes written to disk first:

• node scripts/check-adr-links.mjs :: exit 0, --self-test :: exit 0
• node scripts/check-closing-keyword-parity.mjs :: exit 0, --self-test :: exit 0
• node scripts/check-comment-mask-corpus.mjs :: exit 0
• node scripts/ci/plan-test-core-shards.mjs --self-test :: exit 0
• pnpm --filter @objectstack/lint run check:doc-formula-expressions :: exit 0. The first run exited 3, a prerequisite refusal that measured nothing (formula/lint dist absent). It went green after turbo run build --filter=@objectstack/formula --filter=@objectstack/lint (lock VERDICT command-exit 0).
• pnpm check:ci-filter-parity, check:cross-package-test-inputs, check:doc-authoring, check:driver-memory-census, check:gitlink-declared, check:nul-bytes, check:platform-checklist, check:refd-timer-probe, check:watch-hint-literal :: exit 0 each
• Outside the derivation (the ui frame's standing gate): node scripts/check-sdui-manifest.mjs :: exit 0

dispatch-gates --ran: 16 derived, 16 run, 0 NOT-MEASURED.

Acceptance notes

• No changeset: the diff touches only docs/qa/**, which no package's files[] ships, so the PR carries skip-changeset.
• The quoted strings were read from ui/packages/i18n/src/namespaces/studio/en.ts (lines 3140–3145) and the toast branch in ui/packages/app-shell/src/views/studio-design/StudioDesignSurface.tsx at main efb1620. The toast was not re-driven in a browser.

---

Generated by Claude Code

[objectstack-fleet]

domain:ui#1 · os-tesla · session_01L9qibw2uhdSeHYfnZdAf4V

席位意见:同意。平台清单 studio-authoring.first-run-loop 第 4 步改为引用控制台现在实际显示的发布提示(按数量、无「atomic」、发布对象时附带权限集提示与 Open Access),revision 8 → 9 并追加 history。#27521 余下的第 9 行由 #28519 承接。席位落地。

席位核对

ACCEPT · head 5e94fc038d410a18716d738ca65b3abdd6dce303 · 2026-11-01T05:04Z

1. 形态: 1 file, +4/−3; Fixes #27521; 0 governed paths (check-governed-merges --pr 28526: NOT governed), under 3000 lines. Draft, base main, mergeable_state clean. Label skip-changeset (docs-only, no package touched).
2. 前提实测: claim 6553466756 (surface docs/qa/platform-checklist/areas/studio-authoring.json). Seat read the diff and checked the quoted copy against origin/main: engine.studio.publishedAllCount = "Published {count} items in this package", publishedAllCountOne = "Published {count} item in this package", engine.studio.publishedAccessHint = "Members can use it once a permission set grants them access." (ui/packages/i18n/src/namespaces/studio/en.ts:3142–3145), chosen at StudioDesignSurface.tsx:1895. The step text matches.
3. 门禁 check-runs: 47 on the head, 27 success + 20 skipped, 0 failing. Dev: 16/16 derived gates exit 0 including check:platform-checklist.
4. 契约复审: Clause-②: no, Metadata-protocol: none. None of the four contract faces is touched (docs JSON only), so no record is owed.
5. 范围外去向: row 9 (dismiss label on the Cancel Invitation confirm) goes to spec card Action confirm and parameter dialogs: a declared dismiss label (ActionSchema.cancelText), so "Cancel Invitation" no longer sits beside "Cancel" #28519 (ActionSchema.cancelText). The PR body carries the full row ledger; every other row was already disposed. The file's remaining "atomic" lines (429, 464, 527) describe server publish semantics or quote the audit, are still accurate, and stay.
6. 落地条款: seat runs pr_ready and arms auto-merge on the maintainer's go in this seat's thread, 2026-10-31T23:00Z 「使用 pm-dispatch 技能审核pr后落地」.