@@ -49,6 +49,7 @@ import { BILLING_CALLBACK_OUTCOME } from '@/lib/mothership/generated/billing-pro
 import { BillingRouteOutcome } from '@/lib/mothership/generated/trace-attribute-values-v1'
 import { TraceAttr } from '@/lib/mothership/generated/trace-attributes-v1'
 import { TraceSpan } from '@/lib/mothership/generated/trace-spans-v1'
+import { recordJobRunSpend } from '@/lib/mothership/jobs/budget'
 import { checkInternalApiKey } from '@/lib/mothership/request/http'
 import { withIncomingGoSpan } from '@/lib/mothership/request/otel'
 
@@ -122,6 +123,28 @@ async function readUsageStanding(
   }
 }
 
+/**
+ * Adds a scheduled job run's own spend cap to the payer's standing. A payer over its plan limit
+ * keeps the plan card; otherwise a job run at its cap pauses with the job's card. A cap that
+ * cannot be read answers as the payer's standing alone, like a standing read that fails.
+ */
+async function withJobRunBudget(
+  verdict: BillingUsageVerdict,
+  billingRequestId: string | undefined,
+  cumulativeCostUsd: number
+): Promise<BillingUsageVerdict> {
+  if (verdict.usageExceeded || !billingRequestId) return verdict
+  try {
+    const card = await recordJobRunSpend(billingRequestId, cumulativeCostUsd)
+    return card ? { usageExceeded: true, usageUpgrade: card } : verdict
+  } catch (error) {
+    logger.warn('Job run spend could not be recorded; answering the payer standing alone', {
+      error: toError(error).message,
+    })
+    return verdict
+  }
+}
+
 function getBillingResolution(
   isMarkerlessLegacy: boolean,
   billingAttribution: BillingAttributionSnapshot | undefined
@@ -442,7 +465,11 @@ async function updateCostInner(req: NextRequest, span: Span): Promise<NextRespon
       await checkAndBillOverageThreshold(userId, undefined, { onError: 'throw' })
     }
 
-    const usageVerdict = await readUsageStanding(userId, billingAttribution, accountDecision)
+    const usageVerdict = await withJobRunBudget(
+      await readUsageStanding(userId, billingAttribution, accountDecision),
+      billingRequestId,
+      cost
+    )
     const duration = Date.now() - startTime
 
     // Same-or-lower cumulative than already recorded: nothing new to bill.
 
@@ -108,6 +108,11 @@ vi.mock('@/lib/execution/manual-cancellation', () => ({
 
 vi.mock('@/lib/core/async-jobs', () => asyncJobsMock)
 
+/** The job lane has its own integration suite. */
+vi.mock('@/lib/mothership/jobs/lane', () => ({
+  runJobLane: vi.fn(async () => ({ claimed: 0 })),
+}))
+
 vi.mock('drizzle-orm', () => ({
   and: vi.fn((...conditions: unknown[]) => ({ type: 'and', conditions })),
   eq: vi.fn((field: unknown, value: unknown) => ({ field, value, type: 'eq' })),
 
@@ -50,6 +50,7 @@ import {
   registerManualExecutionAborter,
   unregisterManualExecutionAborter,
 } from '@/lib/execution/manual-cancellation'
+import { runJobLane } from '@/lib/mothership/jobs/lane'
 import {
   buildCarrierReconciledMetadata,
   carrierNotReconciledSql,
@@ -1788,6 +1789,13 @@ export async function runScheduleTick(requestId: string): Promise<ScheduleTickRe
     await Promise.allSettled(schedulePromises)
   }
 
+  try {
+    const { claimed } = await runJobLane()
+    totalSchedules += claimed
+  } catch (error) {
+    logger.error(`[${requestId}] Scheduled job lane failed`, { error: toError(error).message })
+  }
+
   const totalCount = totalSchedules
   const durationMs = Date.now() - tickStart
   logger.info(
 
@@ -523,3 +523,28 @@ describe('CredentialDisplay link tag', () => {
     act(() => root.unmount())
   })
 })
+
+describe('UsageUpgradeDisplay', () => {
+  it("shows a job run's own spend cap without an account billing action", () => {
+    ;(globalThis as { IS_REACT_ACT_ENVIRONMENT?: boolean }).IS_REACT_ACT_ENVIRONMENT = true
+    const container = document.createElement('div')
+    const root = createRoot(container)
+    act(() => {
+      root.render(
+        <SpecialTags
+          segment={{
+            type: 'usage_upgrade',
+            data: {
+              reason: 'job_run_budget',
+              action: 'increase_limit',
+              message: 'This scheduled job run reached its $5.00 spend limit.',
+            },
+          }}
+        />
+      )
+    })
+
+    expect(container.querySelector('a')).toBeNull()
+    act(() => root.unmount())
+  })
+})
@@ -3443,6 +3443,43 @@ function MothershipErrorDisplay({ data }: { data: MothershipErrorTagData }) {
 }
 
 function UsageUpgradeDisplay({ data }: { data: UsageUpgradeTagData }) {
+  return data.reason === 'job_run_budget' ? (
+    <JobRunBudgetDisplay message={data.message} />
+  ) : (
+    <PlanUsageLimitDisplay data={data} />
+  )
+}
+
+/**
+ * A scheduled job's run reached its own spend cap. The cap belongs to the job, not to the
+ * account, so the card offers no billing action: the next run starts with a fresh cap.
+ */
+interface JobRunBudgetDisplayProps {
+  message: string
+}
+
+function JobRunBudgetDisplay({ message }: JobRunBudgetDisplayProps) {
+  return (
+    <InteractionCard
+      title={
+        <span className='inline-flex items-center gap-2'>
+          <TriangleAlert
+            aria-hidden
+            className='size-[14px] shrink-0 text-[var(--badge-amber-text)]'
+          />
+          Run Spend Limit Reached
+        </span>
+      }
+    >
+      <div className='px-2 pb-2'>
+        <p className='text-[var(--text-body)] text-small leading-5'>{message}</p>
+      </div>
+    </InteractionCard>
+  )
+}
+
+/** The payer's plan limit: the card links to where the plan or the limit can change. */
+function PlanUsageLimitDisplay({ data }: { data: UsageUpgradeTagData }) {
   const { data: session } = useSession()
   const hostContext = useOptionalWorkspaceHostContext()
   const organizationContext = useOptionalOrganizationContext()
 
@@ -24,6 +24,7 @@ import { useQueryClient } from '@tanstack/react-query'
 import { defaultRangeExtractor, type Range, useVirtualizer } from '@tanstack/react-virtual'
 import { ComputerUseActivity } from '@/components/computer-use/activity'
 import { SMOOTH_CHASE_RATE } from '@/lib/core/utils/smooth-bottom-chase'
+import type { TurnOrigin } from '@/lib/mothership/generated/protocol'
 import type { WorkspaceFileRecord } from '@/lib/uploads/contexts/workspace'
 import { inter } from '@/app/_styles/fonts/inter/inter'
 import { FindBar } from '@/app/workspace/[workspaceId]/components/find-bar'
@@ -187,24 +188,36 @@ interface UserMessageRowProps {
 }
 
 /**
- * A background-task notification that opened a turn (mothership 21-background-tasks.md
- * §6.4): a muted system chip, never a user bubble — the user did not type it. Shows the
- * outcome line; the provenance header and the agent's own note stay out of the way.
+ * A turn automation opened, never a user bubble — the user did not type it: a muted system chip
+ * with one line. A background-task notification (mothership 21-background-tasks.md §6.4) shows
+ * its outcome line; a scheduled job's run shows the brief's first line (the job and its run).
  */
-const TaskNotificationRow = memo(function TaskNotificationRow({
-  content,
-  rowClassName,
-}: {
+interface AutomatedTurnRowProps {
+  origin: TurnOrigin
   content: string
   rowClassName: string
-}) {
+}
+
+const AutomatedTurnRow = memo(function AutomatedTurnRow({
+  origin,
+  content,
+  rowClassName,
+}: AutomatedTurnRowProps) {
   const lines = content.split('\n').filter((line) => line.length > 0)
-  const outcome = lines.find((line) => line.startsWith('Task ')) ?? lines[lines.length - 1] ?? ''
+  const summary =
+    origin === 'task'
+      ? (lines.find((line) => line.startsWith('Task ')) ?? lines[lines.length - 1] ?? '').replace(
+          /^Task [0-9a-f-]+ /,
+          ''
+        )
+      : (lines[0] ?? '').replace(/^Job: /, '')
   return (
     <div className={cn('flex w-full justify-center', rowClassName)}>
       <div className='max-w-[85%] rounded-full border border-[var(--border)] bg-[var(--bg)] px-3 py-1 text-[12px] text-[var(--text-secondary)]'>
-        <span className='mr-1.5 font-medium text-[var(--text-primary)]'>Background task</span>
-        {outcome.replace(/^Task [0-9a-f-]+ /, '')}
+        <span className='mr-1.5 font-medium text-[var(--text-primary)]'>
+          {origin === 'task' ? 'Background task' : 'Scheduled job'}
+        </span>
+        {summary}
       </div>
     </div>
   )
@@ -930,8 +943,12 @@ export function MothershipChat({
                     style={{ top: virtualItem.start }}
                   >
                     {msg.role === 'user' ? (
-                      interactionPairing.hiddenUserByIndex[index] ? null : msg.origin === 'task' ? (
-                        <TaskNotificationRow content={msg.content} rowClassName={styles.rowGap} />
+                      interactionPairing.hiddenUserByIndex[index] ? null : msg.origin ? (
+                        <AutomatedTurnRow
+                          origin={msg.origin}
+                          content={msg.content}
+                          rowClassName={styles.rowGap}
+                        />
                       ) : (
                         <UserMessageRow
                           content={msg.content}
 
@@ -112,6 +112,18 @@ describe('useChatFind', () => {
     expect(find.current).toMatchObject({ isOpen: false, query: '', count: 0 })
   })
 
+  it('leaves turns automation opened out of the search, whatever opened them', () => {
+    const find = renderFind({
+      messages: [
+        { id: 't', role: 'user', origin: 'task', content: 'Task abc finished: triage done.' },
+        { id: 'j', role: 'user', origin: 'job', content: 'Job: Nightly triage (Run 1)' },
+        assistant('a', 'The triage is done.'),
+      ],
+    })
+    find.search('triage')
+    expect(find.current.count).toBe(1)
+  })
+
   it('counts the same footnoted matches in the index as it highlights in the rendered message', async () => {
     const find = renderFind({
       chatId: 'chat',
 
@@ -163,7 +163,7 @@ export function useChatFind({
     if (term) {
       for (let messageIndex = messages.length - 1; messageIndex >= 0; messageIndex--) {
         const message = messages[messageIndex]
-        if (!message || hiddenUserByIndex[messageIndex] || message.origin === 'task') continue
+        if (!message || hiddenUserByIndex[messageIndex] || message.origin) continue
         const text = getMessageSearchText(message, textCache)
         const occurrences: ChatFindMatch[] = []
         let found = 0
 
@@ -2,7 +2,7 @@ import type { WorkspaceSearchFilters } from '@/lib/api/contracts/knowledge/searc
 import type { MothershipChat } from '@/lib/api/contracts/mothership-chats'
 import type { ManagedMcpConnectorId } from '@/lib/credential-groups/managed-mcp-connectors'
 import type { AssistantSearchLevel } from '@/lib/mothership/generated/assistant'
-import type { ToolActivity } from '@/lib/mothership/generated/protocol'
+import type { ToolActivity, TurnOrigin } from '@/lib/mothership/generated/protocol'
 import type { ChatContext } from '@/stores/panel'
 import type { BrowserTextSelection, TerminalTextSelection } from '@/stores/panel/types'
 
@@ -197,8 +197,8 @@ export interface ChatMessage {
   role: 'user' | 'assistant'
   requestMode?: 'agent' | 'assistant' | 'plan'
   content: string
-  /** "task": a background-task notification opened this turn, not the user (a system chip, not a bubble). */
-  origin?: 'task'
+  /** Automation opened this turn, not the user: a system chip, not a bubble. */
+  origin?: TurnOrigin
   contentBlocks?: ContentBlock[]
   attachments?: ChatMessageAttachment[]
   fileAttachments?: FileAttachmentForApi[]
 
@@ -5,7 +5,11 @@ import {
   usageLimitReadSchema,
   usageLimitUpdateSchema,
 } from '@/lib/billing/application/usage-limit-validation'
-import { BillingCallbackBody, BillingCallbackHeaders } from '@/lib/mothership/generated/billing'
+import {
+  BillingCallbackBody,
+  BillingCallbackHeaders,
+  UsageUpgrade,
+} from '@/lib/mothership/generated/billing'
 
 const booleanQueryParamSchema = z
   .preprocess((value) => {
@@ -315,8 +319,9 @@ export const billingSwitchPlanResponseSchema = z.object({
  * limit writes the tag with this payload verbatim into its durable log.
  */
 export const usageUpgradePayloadSchema = z.object({
-  reason: z.literal('usage_limit'),
-  action: z.enum(['upgrade_plan', 'increase_limit']),
+  /** `job_run_budget`: a scheduled job's run reached its own spend cap, not the payer's plan. */
+  reason: UsageUpgrade.shape.reason,
+  action: UsageUpgrade.shape.action,
   message: z.string(),
 })
 export type UsageUpgradePayload = z.infer<typeof usageUpgradePayloadSchema>