diff --git a/‎.changeset/group-scheduled-work-per-record-ownership.md‎
Lines changed: 132 additions & 0 deletions b/‎.changeset/group-scheduled-work-per-record-ownership.md‎
Lines changed: 132 additions & 0 deletions
diff --git a/‎content/docs/automation/flows.mdx‎
Lines changed: 58 additions & 20 deletions b/‎content/docs/automation/flows.mdx‎
Lines changed: 58 additions & 20 deletions
@@ -0,0 +1,132 @@
+---
+"@objectstack/types": minor
+"@objectstack/spec": minor
+"@objectstack/trigger-schedule": minor
+"@objectstack/metadata-core": minor
+"@objectstack/cli": patch
+---
+
+feat(spec,types,triggers)!: `group` runs package-authored scheduled work without a declaration, owning each run's writes per record (#18378)
+
+<!-- adr-0087: not-required (already-registered schedule-flow-acting-organization-required) This amends the EXISTING semantic entry rather than adding one: same authorable key, same deployment switch, same surface, and the entry predates this diff at the merge base. Nothing is renamed, retired or re-typed — the start node's `config` is an open record (ADR-0018), so every flow that parses today parses byte-identically afterwards and `objectstack migrate meta` has nothing new to rewrite. What moves is the BIND-time accept set (it WIDENS) and the RUN-time organization such a flow's writes carry; the entry's own surface/replacement/reason/acceptanceCriteria each gained their `group` row in this diff. -->
+
+`Clause-②: yes (widening)`
+
+**ADR-0087 disposition — `not-required (already-registered)`, not `registered`.**
+The ledger entry this change belongs to already exists
+(`schedule-flow-acting-organization-required`, entry 18) and predates this diff
+at the merge base, so `registered` would assert a registration this PR did not
+make. The entry's `surface`, `replacement`, `reason` and `acceptanceCriteria`
+each gained their `group` row here, the rejected bootstrap-organization arm
+included — recorded because it is the one a later reader will re-propose.
+
+**Marked breaking (`!`) for the behaviour change, not for a narrowing.** Nothing
+that worked stops working and nothing that was admitted becomes refused — the
+accept set WIDENS in one cell. What earns the banner is the other direction: on a
+`group` deployment with the switch already on, flows that were refused at bind
+now arm and run, so clock-driven work appears where an operator had none. That is
+worth reading before upgrading even though no consumer has to change anything.
+
+## What changes
+
+With `OS_AUTOMATION_SCHEDULED_WORK_ENABLED` on and tenancy posture `group`, a
+time-triggered flow that declares no `config.organization` now **binds and
+runs**, where it was previously refused at bind. The organization its writes
+carry follows the record:
+
+| posture | declaration | a bound run's writes act as |
+|---|---|---|
+| `single` | not read | nothing — the install's one organization resolves beneath each write |
+| `group` | **optional** | declared ⇒ the declaration; undeclared ⇒ **the swept record's own organization** |
+| `isolated` | **required** | the declaration; undeclared ⇒ not armed, unchanged |
+
+A `timeRelative` sweep under `group` reads group-wide — inherent to the posture
+(ADR-0105 D1) — and stamps each run it launches with that record's organization:
+sweep contracts across four plants and each plant's contract yields a run acting
+as that plant, whose notifications reach that plant's inboxes.
+
+## Why this is not a fallback that guesses
+
+It is the order `sys_automation_run` was **already** ruled to use.
+`ObjectStoreSuspendedRunStore` resolves a run's organization as
+`organizationOf(<subject record>) ?? ctx.tenantId` — subject first, acting
+context as the fallback and never the primary. Before this change those two
+halves disagreed under `group`: the history row was stamped from the record while
+the inbox and delivery rows followed an acting context that could not exist
+there, so they were refused while the tick summarised itself as healthy.
+
+⚠️ With one stated exception, because the two halves ask different questions:
+the history row is STAMPED (`tenancy.organizationField` wins there) while the
+run's acting organization is a WALL reading that never consults that key. They
+agree on every object where the two coincide — which is every ordinary object,
+since a declared stamp column is what makes them differ and one shipped object
+declares one (`sys_api_key`, deliberately unwalled). Sweeping that object under
+`group` stamps its history row while the run itself acts as nothing: the correct
+pair of answers, not a residue of the old disagreement, and recorded rather than
+smoothed over.
+
+⛔ A record-less run under `group` that declared nothing still resolves
+**nothing** and is refused at its first tenant-scoped write (`walled-posture`,
+ADR-0112), loudly and by name. The rejected alternative was a fallback to the
+bootstrap organization (`slug='default'`): under a wall that organization is
+minted admin-keyed by the enterprise organizations runtime and may not exist at
+all, and where it does it is whichever organization the platform owner
+registered under — plausibly one plant of many, not the group's head office.
+
+## Upgrading
+
+**Most deployments: nothing to do.** The switch this depends on is OFF by default
+and ships unreleased alongside this change, so the `group`-is-walled behaviour
+being amended has never appeared in a published version — no released consumer
+can be relying on it.
+
+If you run posture `group` **and** turn the switch on, read your boot log: each
+time-triggered flow's bind line now names which of the three shapes it bound as
+("as organization '…'", "with per-record acting organization", or "with NO
+acting organization"). Two things to check:
+
+- A flow you expected to act as ONE organization but which binds per-record is
+  missing its `config.organization`. Add it — declaring still narrows, bounding
+  the sweep's query as well as its identity.
+- A plain `schedule` cron flow that binds "with NO acting organization" has no
+  record to derive one from. If it writes notifications, inbox messages or any
+  other per-organization row, declare `organization` on its start node; the bind
+  line says so, and so does the refusal at the first tick.
+
+## Which organization a record belongs to — the WALL question, not the stamp one
+
+`@objectstack/metadata-core` gains a second face on the record→organization
+resolver, and the split is the point: `resolveRecordOrganizationField` /
+`createRecordOrganizationResolver` answer **"who is this row ABOUT"** (the STAMP
+question, whose `tenancy.organizationField` limb stays pinned to the three
+sanctioned platform-row writers), while the new
+`resolveRecordWallOrganizationField` / `createRecordWallOrganizationResolver`
+answer **"what is this row WALLED by"** — `tenancy.enabled: false` ⇒ nothing,
+then a declared `tenancy.tenantField`, then the kernel's `organization_id`.
+
+The sweep uses the WALL face, because "which organization does this run act as"
+is a question about the wall. ⛔ It never reads `tenancy.organizationField`: that
+key is declared on exactly one shipped object (`sys_api_key`, deliberately
+unwalled, #8287), and reading it here would turn "the audit trail should follow
+this row's own organization even though nothing walls it" into an acting
+identity. A sweep over such an object resolves **nothing** and takes the
+`walled-posture` refusal at its first tenant-scoped write, which is the honest
+answer. Limbs 1 to 4 are one implementation shared by both faces, pinned as
+such, so the half they agree on cannot drift apart.
+
+**API:** `ScheduledWorkPolicy` gains `runOwnership: 'unscoped' | 'per-record' |
+'declared'`, and `requiresActingOrganization` narrows from "any walled posture"
+to `isolated` only. The two are deliberately separate axes: the boolean decides
+whether BIND refuses, `runOwnership` decides what a run that DID bind carries.
+Inside `@objectstack/trigger-schedule`, both triggers share one bind-line
+vocabulary (`describeScheduleRunOwnership`) so they cannot describe one
+deployment differently. ⚠️ That helper is module-level, NOT a package export: it
+is not re-exported from the package barrel, whose own note says an export whose
+only consumers live inside its own package belongs in a non-barrel module. The
+new PUBLIC surface in this change is `ScheduledRunOwnership` and the
+`runOwnership` key on `@objectstack/types`, plus
+`resolveRecordWallOrganizationField` and
+`createRecordWallOrganizationResolver` on `@objectstack/metadata-core` — and
+those four are what put `Clause-②` at `yes`. Nothing existing is renamed or
+re-typed: both stamp-face exports keep their names, their signatures and their
+answers, limb 0 included.
@@ -2078,10 +2078,11 @@ export const renewalReminder: Flow = {
           offsetDays: [60, 30, 7], // — or — withinDays: 30 (negative = overdue lookback)
           filter: { status: 'active' }, // optional, ANDed with the date window
         },
-        // Required under a WALLED tenancy posture, and for a stronger reason
-        // than a plain schedule flow — it bounds the sweep's query as well as
-        // its runs. Not required under `single`. See "The acting organization"
-        // below.
+        // Required under `isolated`, and for a stronger reason than a plain
+        // schedule flow — it bounds the sweep's query as well as its runs.
+        // Optional under `group` (an undeclared sweep reads group-wide and each
+        // run acts as its own record's organization) and not read under
+        // `single`. See "The acting organization" below.
         organization: '<sys_organization.id>',
         // schedule: { type: 'cron', expression: '0 8 * * *' } // optional; defaults to daily 08:00 UTC
       },
@@ -2142,10 +2143,9 @@ it: the caller's session rides into the run and every tenant-scoped write below
 resolves the same organization a normal write would. A **time-triggered** flow
 has no such caller — a job tick carries no identity at all.
 
-Under a **walled** tenancy posture (`group` or `isolated`) that is a question
-only the author can answer, so a `schedule` or `timeRelative` flow there
-**declares the organization it runs as**, on the start node's `config`, beside
-the cadence it scopes:
+Under the **`isolated`** tenancy posture that is a question only the author can
+answer, so a `schedule` or `timeRelative` flow there **declares the organization
+it runs as**, on the start node's `config`, beside the cadence it scopes:
 
 ```typescript
 config: {
@@ -2193,7 +2193,33 @@ scheduled-work switch, and neither is visible from a stack. A lint rule that
 fired on the default posture would be wrong more often than right.
 </Callout>
 
-**Under a wall, a time-triggered flow that declares none is a declaration
+<Callout type="info">
+**Under `group` the declaration is optional, and an undeclared flow still
+runs.** `group` is one legal group over one shared database — group-wide
+visibility and cross-organization workflow are inherent to the shape, not
+violations of it — so a group-level batch job is a capability of the posture. An
+undeclared `timeRelative` sweep there **reads group-wide** and stamps each run it
+launches with **that record's own organization**: sweep a contracts table across
+four plants and each plant's contract produces a run acting as that plant, whose
+notifications land in that plant's inboxes.
+
+That is not a new rule so much as one being made consistent: the
+`sys_automation_run` history row was already stamped from the subject record,
+with the acting context only as a fallback. Before this, the history row and the
+inbox rows of the same run could disagree about who owned it.
+
+Declaring on a `group` flow still works and still **narrows**: the declaration
+bounds the sweep's query as well as its identity, exactly as under `isolated`.
+
+⚠️ A **record-less** flow — a plain `schedule` cron with no sweep — has nothing
+to derive an organization from. Under `group` it binds and runs, but its first
+tenant-scoped write is **refused**, by name, with the remedy. Declare
+`organization` on such a flow if it writes notifications, inbox messages or other
+per-organization rows. The bind line says so at boot rather than leaving it to
+surface at the first tick.
+</Callout>
+
+**Under `isolated`, a time-triggered flow that declares none is a declaration
 error**, refused at bind:
 
 - the trigger logs the reason at `error`, naming the flow;
@@ -2203,15 +2229,19 @@ error**, refused at bind:
   `bound: false`;
 - nothing fires it.
 
-There is deliberately **no fallback** — not the platform organization, not "the
-first row of `sys_organization`", and never the swept record's own
-`organization_id`. Behind a wall, a run that reached every tenant-scoped write
-with nothing to offer would have each of those writes refused one layer below
-anything that summarises the run: the tick reports itself healthy and delivers
-nothing. A wrong `organization_id` is worse still, because it is silently
-authoritative to every report, export and cleanup script that filters by
-organization. ⛔ This is unchanged by the posture split above: `single` **omits**
-the organization, it never invents one.
+There is deliberately **no invented fallback** — not the platform organization,
+not "the first row of `sys_organization`", not the group's bootstrap
+organization. Behind a wall, a run that reached every tenant-scoped write with
+nothing to offer would have each of those writes refused one layer below anything
+that summarises the run: the tick reports itself healthy and delivers nothing. A
+wrong `organization_id` is worse still, because it is silently authoritative to
+every report, export and cleanup script that filters by organization.
+
+⛔ The swept record's own `organization_id` is **not** such an invention, and it
+is used under `group` alone — there the row is the only honest owner available
+and the posture's reads already span the group. Under `single` the organization
+is **omitted**, never filled from the row; under `isolated` an undeclared flow
+never binds in the first place.
 
 **No fan-out.** A single flow belongs to one organization. A sweep wanted in
 several organizations is declared once per organization.
@@ -2220,8 +2250,16 @@ several organizations is declared once per organization.
 The start node's `config` is an open record, so a near-miss spelling —
 `organizationId`, `organization_id`, `orgId`, `org_id`, `tenantId` — parses
 happily and is then ignored. The bind-time refusal names the spelling you
-actually wrote — under a wall, which is the only place the key is required and
-therefore the only place a near-miss is a mistake.
+actually wrote — under `isolated`, which is the only posture where the key is
+required and therefore the only place a near-miss is unambiguously a mistake.
+</Callout>
+
+<Callout type="warn">
+Under `group` a near-miss is **not** reported, because an undeclared flow is a
+legal shape there and the trigger cannot tell "meant to declare, misspelled it"
+from "meant not to declare". The symptom to watch for instead is a sweep whose
+runs act per-record when you expected them all to act as one organization —
+check the bind line, which names which of the two shapes the flow bound as.
 </Callout>
 
 ### Update-triggered flow