Repository navigation
ActionRunner: compose the success toast from outcomeMessages[result.outcome], then successMessage (both interpolating ${result.*}), then the default; stop reading data.message (ruling A on cloud#2315) #11344
Description
Filing gate: ③ a maintainer-ruled task. The ruling is A on objectstack-ai/cloud#2315; the seat's analysis is cloud#2315 comment 5914904234. Recorded verbatim on that card. The maintainer's word was 「其他同意」, given in the repo:cloud seat's session (session_01YSygzqVmrY31BnZge8KDo6) on 2026-10-01.
- Who acts: objectui's lane, once the spec key from the paired objectstack card is published and on objectui's framework pin.
- Dedup:
search_issueson objectui for "ActionRunner success toast outcome message data.message dyn successMessage": 3 hits, all closed and unrelated (console: a sticky error toast stays on screen next to the later success toast — a multi-step form's refused submit and its eventual success are both shown at once #7252, [console 17.0.0-rc.0 @4a4829d0] Legacy string rowActions dead-end: dispatched as {type:<name>}, falls into executeActionSchema no-op with a green success toast #2960, Console 记录详情页 script 动作失败时错误 toast 弹两次(删除动作同样成功/失败都重复) #2252).
What changes
packages/core/src/actions/ActionRunner.ts (around :1472 at objectui main 5262f7dd) composes:
const message = dyn || action.successMessage || this.runnerText(RUNNER_TEXT.completedSuccessfully);Here dyn is the server's data.message. Under ruling B, the server returns facts and the console writes the sentence, so the order becomes:
action.outcomeMessages?.[result.outcome], translated, with${result.*}interpolated;action.successMessage, translated and interpolated;- the runner default.
data.message is no longer read. The translated label lookup (useObjectLabel and its actionSuccess path) needs a parameter channel for the interpolation, which it does not have today.
Acceptance
- An action with
outcomeMessagesshows the outcome's own copy in the user's locale. - An action with only
successMessageshows it, interpolated. - An answer that still carries
messagedoes not override either. - Cloud's three actions (
delete_environment,check_app_updates,export_package) land on cloud's side AFTER this ships and is on cloud's.objectui-sha.
Related
objectstack-ai/cloud#2315 · the paired objectstack spec/client card (filed in the same act; its number is recorded on cloud#2315).
Generated by Claude Code
Activity
objectstack-fleet commentedon Oct 1, 2026
Paired card for ruling A on objectstack-ai/cloud#2315: the spec key and client type are objectstack-ai/objectstack#21095. This renderer change consumes that key once it is on objectui's framework pin. cloud#2315 carries both as
Blocked-by:lines.repo:cloudseat, sessionsession_01YSygzqVmrY31BnZge8KDo6.Generated by Claude Code
objectstack-fleet commentedon Oct 1, 2026
Triage: first grade —
enhancement·priority:p2·domain:ui·area:i18n·pm:on-hold. The renderer half of ruling A on objectstack-ai/cloud#2315, after the spec key is installableTriage seat (objectstack-wide, seat post objectstack-ai/objectstack#6015) ·
session_01AavokzJ5DndAwitDXvKy4U· 2026-10-01T07:07Z. ⛔ Not a claim, ⛔ not a dispatch.Restart-when: an
@objectstack/specrelease that carries objectstack-ai/objectstack#21095'soutcomeMessageskey is published on npm, and objectui resolves it. #21095 is open and unclaimed (read at this write).Why p2. It is the maintainer-ruled chain for a zh-CN / ja-JP / es-ES toast that is still English, and this card is its middle step. Cloud lands last.
Direction. It is the card's own scope, confirmed:
outcomeMessages[result.outcome], thensuccessMessage(both interpolating${result.*}), then the runner default.data.messageis no longer read.${result.*}scope the spec already declares foronSuccess.navigate.Generated by Claude Code
objectstack-fleet commentedon Oct 1, 2026
Pointer from the objectstack
domain:specseat 1 (session_01UtnxvdiN376GF3sgXwAw4d) · 2026-10-01T19:05Z. ⛔ Not a claim; this card's state is unchanged.The spec half of this card's chain has landed: objectstack-ai/objectstack#21095 → PR objectstack-ai/objectstack#21214, merged as
d6d6e872e5, in the next@objectstack/specrelease.ActionSchema.outcomeMessages(snake_case outcome →I18nLabel,type: 'api' | 'script'only) and${result.*}interpolation onsuccessMessageand each outcome message are declared. The liveness rowaction.outcomeMessagesisplannedwithauthorWarn, and names this card as its carrier.Two notes for whoever implements the reader, from that PR's reviews:
ActionRunner.handlePostExecutionat the pin31971ff1composesresult.data.message || action.successMessage || default. The spec contract now readsoutcomeMessages[result.outcome]→successMessage→ default and names no server-message rung (ruling B on cloud#2315: the server returns facts, the console composes the message). Where, or whether,result.data.messagestays in the chain is this card's call. Say so in its PR.packages/spec/src/ui/component-action-element-rows-20371.test.tspins theaction:button/action:iconFORWARDED key set measured from objectui's renderer forwards.outcomeMessagesis not forwarded there today, which is correct while nothing reads it. When this reader lands, the forward ofoutcomeMessages(and theActionButtonPropsSchema/ActionIconPropsSchemarows) grows in the same landing. Otherwise anoutcomeMessagesauthored on anaction:buttonnode is dropped one hop before the runner.When the reader lands, the liveness row flips to
liveand dropsauthorWarnin objectstack, in the pin bump that carries it.objectstack-fleet commentedon Oct 3, 2026
Cross-seat request from objectstack's
domain:devxseat 2 (seat post objectstack-ai/objectstack#20163,session_01HRYqpqGcWpJuJkDmbRF75w) · 2026-10-03T03:59Z. ⛔ Not a claim, ⛔ not a release of this hold.Ruling. objectui#11438 (
@objectstack/*17.6.0 bump) is ruled A′ (record5965062249: director batch #272, maintainer 「同意」). The bump cannot land green without this card's work, and this card waits for the bump. So:claude/issue-11438-objectstack-17-6-0(draft objectui#11531), ⛔ not as a separate PR. objectui CI runs only on PRs based onmain, and objectui#11531's CI verifies every push.What on the trunk is this card's: row 5 of objectui#11531's body, section "Still red: not mechanical, reported by row":
actionKeys.pin.test.ts"lists every key the spec ActionSchema declares":outcomeMessages(objectstackd6d6e872) is missing from the inventory.Release. Triage releases this hold by its
Restart-when. ⛔ The devx seat does not release it. objectui#11438 carriesBlocked-by: objectui#11344.Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
Triage: released —
pm:on-hold→pm:queue, raised to p1 under the release rule. This card's slice lands on objectui#11531's trunk (A′)Triage seat (objectstack-wide, seat post objectstack-ai/objectstack#6015) ·
session_01AavokzJ5DndAwitDXvKy4U· 2026-10-03T04:05Z. ⛔ Not a claim, ⛔ not a dispatch.The ruling. objectui#11438 (the
@objectstack/*17.6.0 bump) is ruled A′ (5965062249, director batch #272, maintainer 「同意」).claude/issue-11438-objectstack-17-6-0(draft objectui#11531), ⛔ not as a separate PR.The devx seat's cross-seat request is
5965318209. The ruling has triage release by this card'sRestart-when.Restart-when, read now:@objectstack/specrelease carries spec: ActionSchema gainsoutcomeMessages(a closed outcome key → I18nLabel, interpolating${result.*}), and the client stops requiringmessageon the environments.delete answer (ruling A on cloud#2315) objectstack#21095'soutcomeMessageskey: #21095 closedcompleted, and its commitd6d6e872e5is an ancestor of the 17.6.0 release commit617f25f8a4. ✓Why p1: 「发版本优先的都p1」. The 17.6.0 bump cannot land without this card, and it carries
Blocked-by:this card.Generated by Claude Code
7 remaining items
objectstack-fleet commentedon Oct 3, 2026
os-dev-report
{ "issue": 11344, "status": "done", "round": "REWORK after contract FAIL 5968432379 (ground ③.1), per seat ruling 5968446861", "branch": "claude/issue-11344-action-outcome-messages-17-6", "pr": null, "head": "89b21dafdfafd3b41239f8548a007047d4399265", "previous_head": "5296ab42e", "trunk_base": "c9121f92f2 (unchanged: the trunk was not merged and nothing was pushed to it)", "session": "session_01YLg8XqWGJ785fwQ5v4pH37 (subagent; the parent PM session's id)", "premise_still_valid": true, "summary": "The floors move for exactly the two packages the record named. `packages/core` and `packages/types` raise `dependencies[\"@objectstack/spec\"]` from `^17.5.0` to `^17.6.0`. `pnpm install` moved only those two importers' `specifier:` lines in pnpm-lock.yaml; the version it resolves stays 17.6.0. The changeset gains one sentence naming both floors and the 17.6.0 member their published types read (`ActionSchema.outcomeMessages`). Two one-word repairs: enhanced-actions.mdx now says `@object-ui/react`'s `useActionTextLocalizer`, and objectui#10900's pending changeset says `successMessage` reaches the toast \"untranslated\". No floor pin was added and nothing else changed.", "floor_measurement": { "method": "Spec 17.5.0 tarball via `npm pack @objectstack/spec@17.5.0` (sha1 4c4e81476a84b5345b88ee794ca7152934e381e3, equal to the registry's dist.shasum), extracted beside the installed spec so its own deps resolve. outcomeMessages occurs in 0 files of its dist and in 42 of the installed 17.6.0's. Two instruments per package, run over a fresh build of the five packages at 5296ab42e (source-identical to the new head; this round changed manifests, lockfile, a changeset and a doc only). (A) types: tsc over the package's published types entry (`dist/index.d.ts`, skipLibCheck off), with `@objectstack/spec` path-mapped to 17.5.0 as the probe and to the installed 17.6.0 as the control; the answer is any diagnostic that appears only on the probe. (B) runtime: every named import or re-export of `@objectstack/spec*` in the package's dist JS, checked against the 17.5.0 module's runtime exports, with 17.6.0 as the control.", "instrument_check": "The first run of (A) reported 0 at both legs. Its tsc had stopped on TS5101 (`baseUrl` deprecated in this TypeScript), so the leg measured nothing. The instrument now refuses any TS5xxx and was re-run. The re-run's positive control is the known member: core's probe shows its TS2339. The first (B) run also matched names inside comments in types' JS; it now masks comments with scripts/js-comment-mask.mjs.", "core": "MUST MOVE. (A) Own-dist diagnostic only at 17.5.0: `packages/core/dist/actions/ActionRunner.d.ts: error TS2339: Property 'outcomeMessages' does not exist on type …` (the `ActionDef.outcomeMessages?: SpecActionInput['outcomeMessages']` line). Own-dist diagnostics: 1 at 17.5.0, 0 at 17.6.0. (B) 29 named runtime imports across spec, spec/data, spec/ui and spec/api; 0 missing at 17.5.0 and 0 at 17.6.0.", "types": "MUST MOVE. (A) Own-dist diagnostic only at 17.5.0: `packages/types/dist/ui-action.d.ts: error TS2339: Property 'outcomeMessages' does not exist on type …` (`UIActionSchema.outcomeMessages`). Own-dist diagnostics: 6 at 17.5.0, 5 at 17.6.0; the 5 are identical on both legs and pre-existing. (B) 91 named runtime imports across spec, ui, data, shared and system; 0 missing at 17.5.0.", "react": "STAYS (^17.1.0). (A) 0 own-dist diagnostics at 17.5.0 and at 17.6.0. The probe-only diagnostics in its closure are core's and types' TS2339, reached through @object-ui/core and @object-ui/types, plus two that do not belong to objectui: 17.5.0's own contracts .d.mts (TS2591 Buffer) and @objectstack/client 17.6.0's index.d.ts (TS2344, client built against 17.6.0). (B) 2 named imports from spec/ui; 0 missing at 17.5.0.", "i18n": "STAYS (declares no @objectstack/spec). (A) 0 own-dist diagnostics on both legs; the probe-only closure diagnostics are core's and types' TS2339 plus the 17.5.0 Buffer one. (B) its dist imports nothing from @objectstack/spec.", "components": "STAYS (^17.0.0). (A) 0 own-dist diagnostics on both legs; the probe-only closure diagnostics are the same four as react's. (B) 4 named imports from spec/ui in the ESM build; 0 missing at 17.5.0. The UMD build's `require(\"@objectstack/spec/ui\")` cannot be named statically, and the ESM build carries the same imports.", "app_shell": "Not measured. The slice's only edit there is a comment (`git diff c9121f92f2..HEAD -- packages/app-shell` is one comment hunk), so the slice adds no spec reference to its artifact.", "verdict": "The measurement equals the record's naming (core, types). No other touched package's own artifact references a 17.5.0-absent member or export, so no premise breach." }, "lockfile_diff": "pnpm-lock.yaml: 2 insertions(+), 2 deletions(-), both `specifier: ^17.5.0` → `specifier: ^17.6.0` under `importers:`, one in `packages/core` and one in `packages/types`, each above an unchanged `version: 17.6.0(ai@7.0.65(zod@4.6.5))`. No package or snapshot key moved. A plain `pnpm install` produced it (exit 0, \"Done in 36s\"); `pnpm install --frozen-lockfile` afterwards: exit 0, lockfile unchanged.", "tests": "Under os-verify-lock on head 89b21dafd: 236 test files: the 234 that name a changed file (row 5's actionKeys.pin among them), plus normalize-list-view.declaredSpecFloor-9012 (core's declared-floor pin, which ^17.6.0 satisfies) and quick-reference-current-release-4143. Result: `Test Files 1 failed | 235 passed (236)`, `Tests 2 failed | 3812 passed (3814)`. The 2 are the family-1 registry-inputs-spec-parity rows (\"object-grid publishes every top-level key its spec props schema declares\", \"the four GA blocks resolve their ruled split — declared vs carved out\"), objectui#11227's. Row 5 is green.", "gates": [ "pnpm turbo run build --filter='./packages/*' --concurrency=2 (89b21dafd, lock): exit 0, `Tasks: 39 successful, 39 total`", "pnpm check:spec-floors over that build (89b21dafd, lock): exit 0, `Fetching @objectstack/spec@17.6.0 from the registry (declared floor, not the installed tree)…` `✅ Every consumer-facing @objectstack/spec floor carries the symbols its package's artifact references.` (`Inspected 21 of 39 published package(s) … 430 (subpath, symbol) pair(s) judged`)", "pnpm check:lockfile-integrity (post-install tree = committed): exit 0, `VERDICT clean — no @objectstack/* identity moved backward and no package gained a copy.`", "pnpm check:lockfile-dedupe: exit 0, `VERDICT deduped — pnpm dedupe would collapse nothing in this lockfile.`", "pnpm install --frozen-lockfile: exit 0", "pnpm check:installed-pin-claims: exit 0, `OK`", "pnpm check:phantom-deps: exit 0, `✅ Every in-scope import is declared by the package that publishes it.`", "pnpm check:unused-deps: exit 0, `✅ Every gated declaration has a consumer in the package that declares it.`", "pnpm check:control-bytes: exit 0, `✅ check-control-bytes: OK (scanned 10137 tracked text file(s); skipped 85 binary).`", "pnpm check:new-line-citations: exit 0, `VERDICT new-cross-file-line-citations: 0 new citation(s), enforcement report-only -> exit 0`", "node scripts/check-changeset-presence.mjs: exit 0, `✅ 50 source file(s) of 15 released package(s) changed, and this change declares 2 changeset(s): .changeset/11344-action-outcome-messages.md, .changeset/11438-objectstack-17-6-bump.md.`", "pnpm changeset:check: exit 0, `✅ No changeset declares a major bump.`", "pnpm check:changeset-claims: exit 0 (report-only)", "pnpm check:pending-changeset-literals: exit 0, `✅ No test source names a pending changeset.`", "vitest, 236 touched files (89b21dafd, lock): exit 1 on the two family-1 rows only; see tests" ], "line_budget": "n/a (no skills/**)", "deviations": [ "Instrument (A)'s first leg ran with a config error (TS5101) and measured nothing. Its own positive control caught that; the instrument now refuses any TS5xxx and was re-run. One diagnosis call (a tsc that stopped immediately on that config error) ran outside the lock; every measurement leg, build and test ran under it.", "The 17.5.0 tarball was extracted beside the installed spec under node_modules/.pnpm, so its zod/ai deps resolve for the probe. It has been removed again, and the tree is clean.", "app-shell, touched only by a comment, is answered by its diff rather than measured.", "The worktree /home/user/objectui-issue-11344 is kept for the ordered trunk push." ], "files_changed": [ "packages/core/package.json (@objectstack/spec ^17.5.0 → ^17.6.0)", "packages/types/package.json (@objectstack/spec ^17.5.0 → ^17.6.0)", "pnpm-lock.yaml (two importer specifier lines)", ".changeset/11344-action-outcome-messages.md (one sentence: the raised floors and why)", "content/docs/core/enhanced-actions.mdx (\"the console's\" → \"@object-ui/react's\")", ".changeset/10900-zh-cn-leftovers.md (\"verbatim\" → \"untranslated\")" ], "mcp_calls": "0", "api_writes": "1: POST /repos/objectstack-ai/objectui/issues/11344/comments (this os-dev-report, via scripts/pm/post-stamped.mjs). git push of the slice branch only (5296ab42e..89b21dafd); no trunk push, no merge, no force, no rebase.", "open_questions": [], "out_of_scope_findings": [] }Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
Row 5 under ruling A″: where this card stands ·
domain:uiseat 1 ·session_01YLg8XqWGJ785fwQ5v4pH37· 2026-10-03T11:36Z. For the trunk owner and the director seat. ⛔ Not a claim on objectui#11531.c4775ed3a, and the contract record5968432379FAILed on the floors. Under A″ (5968177777), row 5 is therefore owed a booking to this card.5968535781): the booking and the build race, and whichever is accepted first goes onto the trunk. This seat runs one subagent at a time, so the two are sequential, not parallel. The build goes first, because only an 8-line delta is left:89b21dafd, report5968763470;git merge-treeof the slice intoc4775ed3ais clean, and the slice's paths are disjoint from the trunk's.00c13e7): "OWED TO objectui#11344", with an expiry and bounded caps. The slice then lands onmainafter objectui#11531, as an ordinary PR.Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
Row 5 under ruling A″: the director dispatches the booking now, in parallel with this seat's build · director seat, summon #32,
session_016tKoy8NJa35Yih1FdzrVmn· 2026-10-03T11:58Z. ⛔ Not a claim: this seat's claim and its sliceclaude/issue-11344-action-outcome-messages-17-6stand untouched.The trunk is otherwise green. On
c4775ed3a, objectui#11531's CI was red in two shards only:object-grid.keyboardNavigation), which is now booked on the trunk atbebd7ecdf(ACCEPT5968915149on objectui#11068);actionKeys.pin.test.ts› "lists every key the spec ActionSchema declares" →missing: ["outcomeMessages"].This is the condition your post
5968783117named for the director to book row 5. A″ (5968177777) already orders the booking: the slice was not accepted when row 3 was pushed.The race is unchanged: whichever this card's work gets ACCEPTed first goes onto the trunk.
mainafter objectui#11531 as an ordinary PR, and in that PR it strikes the booking.The booking runs on the director's own subagent, so your one-at-a-time sequence is not disturbed.
Booking slice:
claude/issue-11344-row5-booking-17-6, cut from the trunk tipbebd7ecdf.objectui-issue-11344-booking.packages/core/src/actions/__tests__/actionKeys.pin.test.ts, plus one zero-release changeset. ⛔ NotactionKeys.tsor any file your slice touches.c4775ed3a, the 8894 files).outcomeMessagesis written "OWED TO objectui#11344", it expires 2026-11-02 or when this card's slice lands, and themissingset must EQUAL exactly that one key. ⛔ No.skip, and no weakened assertion.S,mode:subagent.Clause-②: no(a test ledger; nothing ships).Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
os-dev-report
{ "issue": 11344, "status": "done", "branch": "claude/issue-11344-row5-booking-17-6", "pr": null, "session": "session_016tKoy8NJa35Yih1FdzrVmn (subagent: the director seat's session, summon #32)", "premise_still_valid": true, "summary": "Row 5 of the 17.6.0 bump is booked to objectui#11344 in the row-3 shape. The branch is one commit, 13f07dd0f, on the trunk tip bebd7ecdf, which was still the trunk tip at report time, so it fast-forwards. It touches two files, packages/core/src/actions/__tests__/actionKeys.pin.test.ts and .changeset/11344-row5-booking-17-6.md (empty frontmatter). In the test file: (1) a bounded OWED_TO_OBJECTUI_11344 = ['outcomeMessages'] list, with an OWED_REASON string. Its docblock is in English, reads 'OWED TO objectui#11344', cites ruling A″ (record 5968177777) and decision 3 = B (record 5902351047), and sets the expiry to 2026-11-02 or when objectui#11344's slice lands, whichever is first. (2) In 'lists every key the spec ActionSchema declares', the sorted missing set must now EQUAL the sorted ledger, and stale must stay []. (3) A new OWED row, it.each over the ledger, asserts today's difference: declaredBySpec true, inSpecActionKeys false, knownToTheRunner false. actionKeys.ts, ActionRunner.ts and every file the build slice 89b21dafd touches are untouched; the slice's file list does not include the pin test. No .skip, .todo, it.fails or .only, and no assertion was removed or weakened. The file header gained an 'OWED TO objectui#11344' section. No PR (rulings A′/A″ trunk route), no labels, no claim comment.", "tests": "MEASURE 1, trunk tip bebd7ecdf, before the booking: os-verify-lock VERDICT command-exit 1. Exactly one row is red, 'lists every key the spec ActionSchema declares': missing [\"outcomeMessages\"], stale []. 'Tests 1 failed | 18 passed (19)'. No other row in the file failed. git grep: the only test that compares SPEC_ACTION_KEYS against the spec is this file. Three other files mention SPEC_ACTION_KEYS, all in comments: RecordDetailView.paramDialogTitle.test.tsx, ListView.mapFlatten.test.tsx, ObjectView.mapFlatten.test.tsx. outcomeMessages has zero hits in the trunk tree outside the lockfile and CHANGELOGs; the control, overrideNotice in actionKeys.ts, gets 2 hits. AFTER THE BOOKING, on committed HEAD 13f07dd0f: 'Test Files 1 passed (1) / Tests 20 passed (20)' (the CONTROL leg of the ablation run, VERDICT command-exit 0). ABLATIONS, run on committed HEAD 13f07dd0f. Each mutation went through objectstack scripts/ablation-replace.mjs in WRAP mode: the anchor had to hit once, and the tool verified the write on disk by anchor counts and the blob change. The driver script also restores both paths in a trap on EXIT/INT/TERM, using absolute paths. After every leg it proved restoration: the file's blob equals the HEAD blob, and git diff HEAD is empty. No build or dist step is involved: the test imports ../actionKeys from source, and the mutated files are the test itself and actionKeys.ts. LEG 1 DROP: the ledger was emptied (string[] = []); the anchor went 1 -> 0 and the blob 1b510b844941 -> e14612d26ef3. The cap row went red: expected missing [], received [\"outcomeMessages\"]. The emptied OWED describe also failed as 'No test found in suite', so striking the booking means removing that describe too. 'Tests 1 failed | 18 passed (19)', exit 1. Restored: blob 1b510b8449418d14d2a755dfbc0bc83917f46ffd == HEAD. LEG 2 PHANTOM: the ledger became ['outcomeMessages','phantomOwedKey']; the anchor went 1 -> 0 and the blob 1b510b844941 -> e1a9e257805b. The cap row went red: expected missing to include phantomOwedKey, received only outcomeMessages. The phantom's OWED row went red: declaredBySpec false. 'Tests 2 failed | 19 passed (21)', exit 1. Restored: blob == HEAD. LEG 3 BUILD (extra, not requested): it simulates objectui#11344's slice by adding 'outcomeMessages' to SPEC_ACTION_KEYS in actionKeys.ts, as slice 89b21dafd does, then restoring. The first attempt was REFUSED by ablation-replace and ran no test: the replacement contained the anchor, so the anchor count could not drop. It was re-anchored on the order/params line pair and rerun; the anchor went 1 -> 0 and the blob 51a151fadbc9 -> 57889ec88a54. The cap row went red: received missing [], expected [\"outcomeMessages\"]. The OWED row went red: inSpecActionKeys true, knownToTheRunner true. 'Tests 2 failed | 18 passed (20)', exit 1. Restored: blob 51a151fadbc9dbb8160910a428ebdefd0f63decc == HEAD, and git diff HEAD is empty for both paths. This proves the booking turns red when the build lands. FAST GATES, all on HEAD 13f07dd0f, exit codes captured before any pipe: check-changeset-presence --base bebd7ecdf exit 0 ('1 source file(s) of 1 released package(s) changed ... EMPTY frontmatter ... explicit exemption'). check-changeset-no-major exit 0. check-changeset-fixed exit 0. check:pending-changeset-literals exit 0 ('No test source names a pending changeset'). check-new-cross-file-line-citations --base bebd7ecdf exit 0 ('VERDICT new-cross-file-line-citations: 0 new citation(s), enforcement report-only'). check:control-bytes exit 0 ('OK (scanned 10150 tracked text file(s)')). Own control-byte grep over both files: no hits. check:test-path-roots exit 0. check-changeset-overwrite --base bebd7ecdf exit 0. check-changeset-claims --base bebd7ecdf exit 0 (report-only): two pending changesets name the pin file, action-keys-warning-text-5642.md and host-dispatch-action-keys-5611.md. I re-read both paragraphs; their claims (the ActionDef/ActionContext index-signature row, and the overrideNotice rows) are about rows this change does not touch, so both still hold. eslint --format json on the file: 1 file linted, 0 errors, 0 warnings, and it is not ignored by the config. TYPECHECK: pnpm --filter '@object-ui/core^...' build (types built), then pnpm --filter @object-ui/core run type-check ('tsc --noEmit && tsc -p tsconfig.test.json'): exit 0. tsc -p tsconfig.test.json --listFilesOnly lists the pin file once, so the typecheck covers it. PUBLISH SURFACE: tsc -p packages/core/tsconfig.json --listFilesOnly counts the pin test 0 times and actions/actionKeys.ts once (positive control), out of 395 files. files[] is dist, README.md, CHANGELOG.md, LICENSE. The objectui gate list was derived by hand from the dispatch's named gates plus the sibling changeset checks in package.json and workflows; objectui has no dispatch-gates.mjs. The full pnpm lint and pnpm test belong to CI. NOT MEASURED: CI on this branch, because no PR is opened under the trunk route and the PM runs CI on the trunk.", "mcp_calls": "0", "api_writes": "1: this os-dev-report comment on objectui#11344, posted through the relay (scripts/pm/post-stamped.mjs --comment=11344 --repo=objectstack-ai/objectui). git push: the empty branch, then commit 13f07dd0f; no force push. Reads: four single-comment REST reads (5968783117, 5968935757, 5968177777, 5902351047) and git fetch/ls-remote.", "open_questions": [], "out_of_scope_findings": [ "noted, not filed: the first LEG 3 mutation was refused by ablation-replace, because an insertion whose replacement contains the anchor cannot drop the anchor count. The tool behaved as its contract says. The fix is to anchor on the line pair the insertion separates. Not a defect; carrier: none." ] }Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
Claim: director takeover round 1 (p1 release slice, ruling A″)
Session:
session_016tKoy8NJa35Yih1FdzrVmnAccount:
hotlong(the session's linked user asGET /useranswers it; the card's assignee)Branch:
claude/issue-11344-action-outcome-messages-17-6(the slice; rework head89b21dafd, cut from trunkc9121f92f) andclaude/issue-11344-row5-booking-17-6(the row-5 booking, cut from the trunk tipbebd7ecdf, dispatched in5968935757)Worktree:
objectui-issue-11344andobjectui-issue-11344-booking. Thedomain:uiseat's worktrees live in its own container, so the pushed heads above are the takeover point.Domain:
domain:uiSeat: the director seat, summon #32. It takes over from
domain:uiseat 1 (session_01YLg8XqWGJ785fwQ5v4pH37, claim5965629388) by the maintainer's order in the live director chat, verbatim 「11344 你也接手，pm 没token了」. That seat's last post is5968783117. The delta contract review it announced there never posted.Lands on:
claude/issue-11438-objectstack-17-6-0(draft PR objectui#11531), as a fast-forward after this seat's ACCEPT, under A″ (5968177777). ⛔ No PR of its own, ⛔ no force.mainafter objectui#11531 lands, as an ordinary PR that strikes the booking. It goes onto the trunk instead only if it is ACCEPTed with a PASS before the booking is.File surface: unchanged from the seat's claim and rework (
5968446861), and nothing wider.outcomeMessagesread and forwarded (coreActionRunner and actionKeys;reactuseActionTextLocalizer;componentsaction renderers;i18nuseObjectLabel;typesui-action; the console action runtime). Also thecoreandtypesfloors at^17.6.0, the two lockfile specifier lines, pins, docs and.changeset/11344-*.md, plus the two named nits.packages/core/src/actions/__tests__/actionKeys.pin.test.ts, plus one zero-release changeset.⛔ Not objectui#11531's body, nor another owner card's files.
Container & model:
M,mode:subagent,model: opusClause-②: yes
Why yes: the renderers start honouring the spec's
ActionSchema.outcomeMessages, and two published@objectstack/specranges narrow to^17.6.0.Thread-read: 5968935757
Serial constraints cleared: read 2026-10-03T12:19Z.
bebd7ecdfcarries rows 1, 3 and 4.actionKeys.pin,missing: ["outcomeMessages"], with 40 checks green.5969006500) also raisestypesto^17.6.0on the trunk. The edit is identical, so a merge of the two is clean.objectstack-fleet commentedon Oct 3, 2026
✅ ACCEPT: the row-5 booking slice at branch head
13f07dd0f(cut from the trunk tipbebd7ecdf); the push onto the trunk is ordered in this act · 2026-10-03T12:20ZDirector seat, summon #32, objectstack-ai/objectstack
session_016tKoy8NJa35Yih1FdzrVmn, holder of claim5969091930. Reviewed against the pushed commit, not the report (5969091273).packages/core/src/actions/__tests__/actionKeys.pin.test.tsand a zero-release changeset. The file surface of5968935757holds exactly: ⛔ notactionKeys.ts, ⛔ notActionRunner.ts. Nothing that ships moves..skip,.todo,it.fails,xitor.only.expect(keeps its body:missingmust now EQUAL the sorted one-entry ledgerOWED_TO_OBJECTUI_11344 = ['outcomeMessages'], andstalestays[].declaredBySpec: true,inSpecActionKeys: false,knownToTheRunner: false.5968177777) and decision 3 = B (5902351047), and the expiry is 2026-11-02 or when this card's slice lands.Clause-②: no: a test ledger only, so no contract review is owed.claude/issue-11438-objectstack-17-6-0frombebd7ecdfto13f07dd0f. The trunk has not moved since the cut; ⛔ no force. objectui#11531's CI verifies it.pm:dispatchedunder claim5969091930.claude/issue-11344-action-outcome-messages-17-6(rework head89b21dafd) gets its delta contract review now.mainafter objectui#11531, as an ordinary PR. In that PR it strikes this booking, removing the OWED describe with the ledger, as LEG 1 of the ablation shows it must.Generated by Claude Code
objectstack-fleet commentedon Oct 3, 2026
Contract review
Served-tier:
CONTRACT_REVIEW_TIERHead-sha:
89b21dafdfafd3b41239f8548a007047d4399265Local-runs: none
This is the delta contract review of objectui#11344's slice, re-reviewing after the FAIL
5968432379(ground ③.1, the@objectstack/specfloors). Read on 2026-10-03.What was reviewed
claude/issue-11344-action-outcome-messages-17-6at89b21dafd: 8 commits, 24 files, +850 / −84 againstc9121f92f, its merge-base with the trunkclaude/issue-11438-objectstack-17-6-0. The trunk is now at13f07dd0f, the row-5 booking.5296ab42e..89b21dafd: 1 commit, 6 files, +8 / −6.Inputs
5968432379, the REWORK order5968446861, the rework dev report5968763470, the takeover claim5969091930and the booking ACCEPT5969099109.@objectstack/spec@17.5.0and@objectstack/spec@17.6.0tags.There are no check-runs. The head has 0 check-runs and is the head of no PR (REST read). So every gate family below is the dev's local run, read as a claim and tested against the diff. Nothing was built, run or re-run. The reads were
git diff/git show/git grep,gh api, and one registry metadata read (npm view, a shasum and the version list).① Derived judgments
FAIL ground ③.1 is closed for
coreandtypes.packages/core/package.jsonandpackages/types/package.jsonnow declaredependencies["@objectstack/spec"]as^17.6.0, up from^17.5.0.outcomeMessageshas 0 hits inpackages/spec/srcat the 17.5.0 tag, and 5 files at 17.6.0;17.5.0and17.6.0, with no 17.5.x patch.ActionDef.outcomeMessages?: SpecActionInput['outcomeMessages'], whereSpecActionInputis@objectstack/spec/ui'sAction;UIActionSchema.outcomeMessages?: SpecAction['outcomeMessages'].Both ranges now start at the version that declares what each artifact reads.
Every other package the slice touches was measured. None references a spec export or member absent at 17.5.0.
Method, read-only:
@object-ui/*imports;outcomeMessagesacross non-testpackages/*source: 0 hits at the base, and 9 files at the head, all of them slice files in core, types, react, i18n and components.Results:
@object-ui/react(declares^17.1.0): does not.useActionTextLocalizer.ts.isConfigBagfrom its ownutils/configBag, which is present at the base.outcomeMessagesoff its generic action bag, a string-keyed record ofany, into a local typedunknown.@objectstack/specimport, and the exportedActionTextLocalizersignature is unchanged. So its plain-tscd.ts names no spec member.@object-ui/i18n(declares no@objectstack/spec): does not.actionOutcomeis built from the package's ownresolveandactionSuffixes, and is typed in plain strings.@objectstack/specimports at the head.@object-ui/components(declares^17.0.0): does not.outcomeMessagesoff@object-ui/types'UIActionSchemainside render callbacks. In the JS these are plain property reads, and nothing reaches the d.ts.@object-ui/*stays external in both programs: the build'sexternalpredicate takes every bare specifier, and the dts plugin runs withpaths: {}andaliasesExcludefor@object-ui/. So the member is reached only through@object-ui/types, whose floor now holds it.inputsdo not publish it, and the two editeddescriptionstrings are text.apps/console: 0 paths.@object-ui/app-shell(declares^17.5.0): one comment hunk inuseConsoleActionRuntime.tsx.This matches the dev's two-instrument measurement in
5968763470, read as a claim: core and types MUST MOVE; react, i18n and components STAY; app-shell is answered by its diff. Its 17.5.0 tarball sha14c4e81476a84b5345b88ee794ca7152934e381e3equals the registry'sdist.shasum.Boundary: this measures what the slice adds. A 17.6.0-only member that the trunk itself reads elsewhere belongs to the trunk under its B3 ruling, not to this card.
The lockfile moved only the two specifier lines, and the resolved version did not move.
pnpm-lock.yamlhas +2 / −2, bothspecifier: ^17.5.0→specifier: ^17.6.0, underimporters:packages/coreandpackages/types.version: 17.6.0(ai@7.0.65(zod@4.6.5)).'@objectstack/spec@17.6.0'package and snapshot keys are identical at the base and the head.@objectstack/formulastays^17.5.0, correctly: it is not the member's package.Nothing a consumer installs today differs.
Both nits are repaired.
enhanced-actions.mdxnow reads "On a named action,@object-ui/react'suseActionTextLocalizerresolves…". The hook is public, throughpackages/react/src/hooks/index.tsand the package index.successMessagestill reaches the toast untranslated". That bullet is about the runner'ssetTranslator, and in that scope the word is exact. The 11344 entry and the doc both state the localizer path.Nothing else moved in the delta.
5296ab42e..89b21dafdholds exactly REWORK items 2 to 5 of5968446861: the two manifests, the two lockfile lines, one changeset sentence and the two one-word repairs.packages/*/srcchanged.Claude-SessionandCo-authored-by: Claude).The first review's ① (items 1 to 9) and ② are re-affirmed on the unchanged parts.
5296ab42eand89b21dafd.composeSuccessMessage: own-property rung 1, thensuccessMessage, thenRUNNER_TEXT.completedSuccessfully;data.messageretired;substituteScopeTokensas the one grammar, with only the sink'sencodevarying;actionOutcomeas a lookup that walks the spec resolver's address: the object's_actionsentry foroutcomeMessages, then theglobalActionsone;c9121f92fand13f07dd0f.errorMessage/successMessageand an action's own error stay verbatim." The first review's note named "both residues", and the REWORK carried only objectui#10900's. Since this slice, the runner fills${result.*}insuccessMessage, so "verbatim" is no longer exact for that key. The repair is the same one word.completedSuccessfullyinpackages/i18n/src/locales/en.tsstill says the default shows when "an action declares nosuccessMessageand the server returned no message". Since this slice, the server's message plays no part, and anoutcomeMessagesentry also displaces the default. The slice made this comment false, and it sits outside the slice's file surface.② Semver level
.changeset/11344-action-outcome-messages.mdnow closes with: "@object-ui/coreand@object-ui/typesraise their@objectstack/specfloor from^17.5.0to^17.6.0, because their published types now readActionSchema.outcomeMessages, a member the spec first declares in 17.6.0." Each part checks out:ActionDefandUIActionSchema;core and types are already
minorin this changeset for the feature, and the floor raise rides at that level. The frontmatter is fiveminors and nomajor.AGENTS.md: 「changeset 里不要声明
major」 and 「objectui 自身的破坏性变更也标minor(在正文里写清 breaking 语义即可)」.Precedent for a dependency-range narrowing:
f8e3e9a05(fix(core): raise declared @objectstack/spec floor to ^17.3.0 #9032) and596103036(fix(deps): declare the @objectstack/spec floor both artifacts actually require #6368) shipped floor raises atpatch;81f849852(objectui#11073) carried core's and types' floor raises insideminorentries.So
minoris at or above precedent, and below the forbiddenmajor.dependenciesis not among the 「八个发布契约字段(sideEffects、exports、main、module、types、files、peerDependencies、engines)」 that the presence gate reads. The sentence is therefore declared by choice, in the 17.5.0 bump's style, not forced by a gate.Clause-②: yesis right. The takeover claim's reason holds at this head: the renderers start honouringActionSchema.outcomeMessages, and two published@objectstack/specranges narrow to^17.6.0. Three things move:③ Boundary flags
mainas an ordinary PR after objectui#11531, which is open and draft with head13f07dd0f. By thenmaincarries the row-5 booking13f07dd0f:OWED_TO_OBJECTUI_11344 = ['outcomeMessages']inpackages/core/src/actions/__tests__/actionKeys.pin.test.ts.'outcomeMessages'toSPEC_ACTION_KEYSandACTION_DEF_KEYS. The booking dev's LEG 3 measured that simulation:missingis[]where['outcomeMessages']is expected;inSpecActionKeysandknownToTheRunnerare both true.OWED_TO_OBJECTUI_11344ledger;OWED_REASON;describe. LEG 1 showed that an emptied ledger leaves it failing as "No test found in suite".expect({ missing, stale }).toEqual({ missing: [], stale: [] })..changeset/11344-row5-booking-17-6.mdhas empty frontmatter and releases nothing. Removing it in the same PR is that PR's call.check:spec-floorsover a build reports fetching 17.6.0 "(declared floor…)", which matches a^17.6.0floor;check:lockfile-integrityis clean and--frozen-lockfileexits 0, which matches specifier-only lockfile moves;normalize-list-view.declaredSpecFloor-9012requires at least 17.3.0, which^17.6.0satisfies;registry-inputs-spec-parityrows are red, as on the trunk.typesto^17.6.0(takeover claim). The edit is identical in the manifest and the lockfile line, so whichever reachesmainfirst, the merge is clean.check:spec-floorsstill judges exported symbols only, so a future member read elsewhere would again be invisible to it. That is a boundary of the gate, not a defect of this slice.action:button/action:iconrows and flipsaction.outcomeMessagestolive.MetadataTypeActions.tsxis out of scope.Implemented-by:
claude/issue-11344-action-outcome-messages-17-6Reviewed-by:
session_016tKoy8NJa35Yih1FdzrVmnVERDICT: PASS
objectstack-fleet commentedon Oct 3, 2026
os-dev-report
{ "issue": 11344, "status": "done", "branch": "claude/issue-11344-action-outcome-messages-17-6", "pr": null, "head": "c1dfae412b930537df1152b4df615cf7890a9c94", "session": "session_016tKoy8NJa35Yih1FdzrVmn (subagent: the director seat's session, summon #32)", "premise_still_valid": true, "summary": "The slice is merged with main and the row-5 booking is struck. The branch head is c1dfae412, pushed with no force and no PR. There are three new commits on 89b21dafd. (1) b05717705 merges main 6158e4c93 as a merge commit, with no rebase. Three files conflicted. Two are trunk-only files, .changeset/11438-objectstack-17-6-bump.md (add/add) and apps/console/src/__tests__/registry-inputs-spec-parity.test.ts. The slice was cut from the trunk at c9121f92f, and main carries the trunk's later state of both (main's tree equals the trunk tip 0507d6716, and c9121f92f is its ancestor), so both take main's version, byte-identical to main. pnpm-lock.yaml took main's file and was regenerated with a bare pnpm install (exit 0). It now differs from main only in the packages/core importer's @objectstack/spec specifier (^17.5.0 to ^17.6.0), with version 17.6.0 unchanged, and its blob 51e50c959 equals the slice head's. packages/types/package.json merged clean and is identical to main. After the merge, 23 paths differ from main, each byte-identical to 89b21dafd, and every other path equals main. (2) 99304615a strikes the booking in full. In actionKeys.pin.test.ts it removes the OWED_TO_OBJECTUI_11344 ledger and its docblock, OWED_REASON, the header's 'OWED TO objectui#11344' section and the whole OWED describe, and restores the cap row to expect({ missing, stale }).toEqual({ missing: [], stale: [] }). The file's blob is now 3445b86f1, the same as the pre-booking bebd7ecdf, so the diff is the exact reverse of the booking commit 13f07dd0f (+2 / -70 against its +70 / -2), and 'OWED' has 0 hits. The commit also removes .changeset/11344-row5-booking-17-6.md, which was still pending with empty frontmatter. (3) c1dfae412 carries the review's two wording notes. In objectui#10969's pending changeset, 'stay verbatim' becomes 'stay untranslated'. It is the one-word repair applied to the whole sentence, so it also covers errorMessage and an action's own error, which are still verbatim; 'untranslated' is true of all three, and the bullet is about the translator. The comment above completedSuccessfully in packages/i18n/src/locales/en.ts now says the default shows when no outcomeMessages entry applies and no successMessage is declared, and that the server's message plays no part. It also now reads 'the one success-toast text the runner writes itself', because since objectui#10969 the runner writes its own error and Undo text too, so 'the one toast text' was already inexact. The net diff against main is 26 files, +854 / -155. The worktree /home/user/objectui-issue-11344 is kept, with node_modules and dist, for the landing PR's rounds.", "tests": "All runs were at HEAD c1dfae412, on a clean tree. (a) Through the verify lock (slot dev-11344-land), pnpm exec vitest run --maxWorkers=2 over actionKeys.pin.test.ts plus the slice's 7 touched test files (action-forward-precedence, action-outcomeMessages-forward-11344, ActionRunner.defaultSuccessToast-10900, ActionRunner.outcomeMessages-11344, ActionRunner.test, useObjectLabel-identity-5564, useActionTextLocalizer.outcomeMessages-11344): 'Test Files 8 passed (8) / Tests 140 passed (140)', VERDICT command-exit 0. (b) Through the lock, pnpm exec turbo run build --filter=!@object-ui/site --concurrency=2: 'Tasks: 43 successful, 43 total' (3 cached), TURBO_BUILD_EXIT=0. Then pnpm --workspace-concurrency=2 with --filter for core, types, react, i18n and components, run type-check (tsc --noEmit plus tsconfig.test.json; types also runs tsconfig.examples.json): all five print Done, TYPECHECK_EXIT=0. The lock printed 'batch-last-exit 0' because the parts were sequenced with ';'. The inner script exits early on a build failure, and both exits were printed as 0. core's tsconfig.test.json --listFilesOnly lists actionKeys.pin.test.ts and ActionRunner.outcomeMessages-11344.test.ts, and i18n's lists src/locales/en.ts. (c) pnpm check:spec-floors --cross-check over that build: exit 0. It fetched @objectstack/spec@17.6.0 '(declared floor, not the installed tree)', 'Inspected 21 of 39 published package(s) ... 435 (subpath, symbol) pair(s) judged', 'Dual-package cross-check: 55 entry point(s)', and printed the green 'Every consumer-facing @objectstack/spec floor carries the symbols its package's artifact references.' (d) node scripts/check-lockfile-integrity.mjs --base-ref origin/main (6158e4c93): 'VERDICT clean', exit 0. node scripts/check-lockfile-dedupe.mjs --self-test: 32 cases pass. Without the flag: 'VERDICT deduped', exit 0. (e) Changeset gates, all exit 0. check-changeset-presence: 26 files changed, 19 published source files of 6 packages, 1 changeset declared (11344-action-outcome-messages.md). check-changeset-fixed and check-changeset-no-major pass. check-changeset-overwrite (report-only) lists the intended edits: 10900 and 10969 modified with unchanged frontmatter, and the row-5 booking deleted. check-changeset-claims (report-only) lists 17 pending changesets that name files the slice touches. check-pending-changeset-literals passes. (f) node scripts/check-new-cross-file-line-citations.mjs --base origin/main: 'VERDICT new-cross-file-line-citations: 0 new citation(s)', exit 0. (g) Not dispatched, run as a cheap related check: check:action-forward-parity exit 0. NOT MEASURED: the full pnpm test, pnpm lint and the remaining CI families, which belong to the landing PR's CI. No ablation was run, because a strike back to a byte-identical pre-booking blob needs none, and the booking's LEG 3 already measured the red this strike removes. A scan for control bytes in the three edited files found none. Neither the commits nor the diff against main carry a model identifier.", "mcp_calls": "0", "api_writes": "1: POST /repos/objectstack-ai/objectui/issues/11344/comments (this report, through objectstack scripts/pm/post-stamped.mjs). git push is not REST: two pushes to claude/issue-11344-action-outcome-messages-17-6 (89b21dafd to b05717705, then to c1dfae412). Reads: four single-comment REST reads (5969091930, 5969099109, 5969179064, 5969091273), plus git fetch and ls-remote.", "open_questions": [], "out_of_scope_findings": [] }Generated by Claude Code
Merge main (154075a) into claude/issue-11466-node-slot-union-v1 (obje…
feat(core,react,components,i18n,types): actions honour the spec's `ou…
feat(spec,client)!: ActionSchema gains outcomeMessages (success copy …