test: remove seven low-value cases and repair behavioral assertions (#33453)

* test: remove low-value cases and repair behavioral assertions

* test: prune repaired wall-clock finding from flaky-test baseline

---------

Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>

config/flaky-test-pattern-baseline.txt

@@ -1051,7 +1051,6 @@
 3 test-no-wall-clock:src/renderer/src/components/activity/activity-auto-mark-read-loop.react185.test.tsx
 2 test-no-wall-clock:src/renderer/src/components/artifacts/artifact-list-windowing.test.tsx
 2 test-no-wall-clock:src/renderer/src/components/artifacts/ArtifactCollection.test.tsx
-1 test-no-wall-clock:src/renderer/src/components/automations/automation-list-search-rows.test.ts
 7 test-no-wall-clock:src/renderer/src/components/automations/AutomationListLocalRows.test.tsx
 1 test-no-wall-clock:src/renderer/src/components/automations/use-automation-list-search.test.tsx
 1 test-no-wall-clock:src/renderer/src/components/browser-pane/assemble-chrome/browser-workspace-pane.adoption-chrome.test.tsx

src/main/agent-launch/startup-agent-worktree-create.test.ts

@@ -82,7 +82,7 @@ const FULL: RuntimeManagedWorktreeCreateArgs = {
 }
 
 describe('createWorktreeWithStartupAgent', () => {
-  it('starts a terminal agent with the prompt folded by the create, whatever the chat default', async () => {
+  it('forwards the startup agent and prompt to the create with the chat default on', async () => {
     const { runtime, created, create } = harness()
     const prompt = 'x'.repeat(3000)
     const args = {
@@ -100,7 +100,7 @@ describe('createWorktreeWithStartupAgent', () => {
     expect(runtime.getStructuredAgentSessionCreateSupport).not.toHaveBeenCalled()
   })
 
-  it('hands a post-start agent its prompt through the create, which sends it once the agent is up', async () => {
+  it('forwards a post-start agent and prompt to the create', async () => {
     const { runtime, create } = harness()
     const args = { ...FULL, startupAgent: 'aider' as const, startupPrompt: 'fix the bug' }
 
@@ -109,7 +109,7 @@ describe('createWorktreeWithStartupAgent', () => {
     expect(runtime.createManagedWorktree).toHaveBeenCalledWith(args)
   })
 
-  it('treats a blank prompt as none, which the create launches bare either way', async () => {
+  it('omits a blank startup prompt from the create request', async () => {
     for (const startupPrompt of ['', '  \n']) {
       const { runtime, create } = harness()
 
@@ -119,7 +119,7 @@ describe('createWorktreeWithStartupAgent', () => {
     }
   })
 
-  it('starts a linked draft through `startupDraft`, unsent, with the agent the draft resolves', async () => {
+  it('forwards a linked draft through `startupDraft` with its resolved agent', async () => {
     const { runtime, create } = harness({ draftAgent: 'codex' })
     const { createdWithAgent: _requested, ...request } = FULL
     const draft = 'https://github.com/o/r/issues/1'

src/main/agent-trust-presets.test.ts

@@ -84,7 +84,7 @@ describe('markCursorWorkspaceTrusted', () => {
       const trustFile = join(projectsDir, slugDirs[0], '.workspace-trusted')
       expect(existsSync(trustFile)).toBe(true)
       const payload = JSON.parse(readFileSync(trustFile, 'utf-8'))
-      expect(payload.workspacePath).toBeTruthy()
+      expect(payload.workspacePath).toBe(realpathSync.native(workspace))
       expect(typeof payload.trustedAt).toBe('string')
     } finally {
       rmSync(workspace, { recursive: true, force: true })
@@ -152,7 +152,7 @@ describe('markCopilotFolderTrusted', () => {
       const parsed = JSON.parse(readFileSync(configPath, 'utf-8'))
       expect(Array.isArray(parsed.trustedFolders)).toBe(true)
       expect(parsed.trustedFolders.length).toBe(1)
-      expect(typeof parsed.trustedFolders[0]).toBe('string')
+      expect(parsed.trustedFolders[0]).toBe(realpathSync.native(workspace))
     } finally {
       rmSync(workspace, { recursive: true, force: true })
     }

src/main/ai-vault-search/session-search-file-write.test.ts

@@ -72,7 +72,7 @@ function counts(db: SyncDatabase): Record<string, number> {
   }
 }
 
-it('writes a whole read in one transaction', () => {
+it('publishes all rows from a completed whole read', () => {
   replayTranscriptRead({ messages: userMessages('needle text', 300) })
 
   const after = counts(index.db)

src/main/ai-vault-search/session-search-index-generation.test.ts

@@ -154,7 +154,7 @@ it('moves the generation when a purge reclaims rows nothing can reach', async ()
   }
 })
 
-it("leaves the generation alone when a replace swaps a session's own rows", async () => {
+it('leaves the generation alone when messages attached to live sessions are deleted', async () => {
   // Deleting messages attached to live sessions must not fire the orphan-reclamation trigger.
   const root = await tempRoot()
   const path = join(root, 'index.sqlite')
@@ -213,7 +213,7 @@ it('keeps the generation across a reopen, because the bump rides its own commit'
   }
 })
 
-it('fences a reader against a writer it does not share a process with', async () => {
+it('fences a reader against another database handle moving the index', async () => {
   // A process-local generation would miss another process’s writes and admit stale cursors.
   const root = await tempRoot()
   const path = join(root, 'index.sqlite')

src/main/ai-vault-search/session-search-message-rows.test.ts

@@ -132,7 +132,7 @@ it('cuts at punctuation when the window holds no whitespace at all', async () =>
   }
 })
 
-it('keeps a 9,000-character identifier whole rather than cutting at its underscores', () => {
+it('hard-cuts a 9,000-character identifier without treating underscores as boundaries', () => {
   // Underscores are token characters, so this oversized identifier requires a hard cut.
   const text = 'ab_'.repeat(3000)
   expect(text.length).toBe(9000)

src/renderer/src/components/automations/automation-create-destination.test.ts

@@ -20,7 +20,7 @@ function entry(overrides: Partial<AutomationHostCatalogEntry> = {}): AutomationH
   return {
     stableRef: { authority: { kind: 'desktop' }, selector: { kind: 'self' } },
     owner: { authority: { kind: 'desktop' }, selector: { kind: 'self' } },
-    stableKey: 'desktop:self',
+    stableKey: 'host:desktop:self',
     label: 'This computer',
     authorityLabel: 'Desktop',
     kind: 'self',
@@ -39,7 +39,7 @@ function sshEntry(targetGeneration: number): AutomationHostCatalogEntry {
       authority: { kind: 'desktop' },
       selector: { kind: 'ssh', targetId: 'box', targetGeneration }
     },
-    stableKey: 'desktop:ssh:box',
+    stableKey: 'host:desktop:ssh:box',
     kind: 'ssh'
   })
 }
@@ -82,13 +82,13 @@ describe('create destination resolution', () => {
   it('prefers the explicit selection and only then the active workspace', () => {
     const entries = [entry(), sshEntry(3)]
     expect(
-      preselectAutomationCreateHost(entries, 'desktop:ssh:box', 'desktop:self')?.stableKey
-    ).toBe('desktop:ssh:box')
-    expect(preselectAutomationCreateHost(entries, null, 'desktop:self')?.stableKey).toBe(
-      'desktop:self'
+      preselectAutomationCreateHost(entries, 'host:desktop:ssh:box', 'host:desktop:self')?.stableKey
+    ).toBe('host:desktop:ssh:box')
+    expect(preselectAutomationCreateHost(entries, null, 'host:desktop:self')?.stableKey).toBe(
+      'host:desktop:self'
     )
     // An unresolvable workspace host leaves the choice open rather than picking one.
-    expect(preselectAutomationCreateHost(entries, null, 'desktop:ssh:gone')).toBeNull()
+    expect(preselectAutomationCreateHost(entries, null, 'host:desktop:ssh:gone')).toBeNull()
     expect(preselectAutomationCreateHost(entries, null, null)).toBeNull()
   })
 })
@@ -172,14 +172,14 @@ describe('offered create hosts', () => {
   it('does not blame the server for hosts another repair or none would fix', () => {
     // Unverified since disconnect: the repair is a reconnect, not an update.
     const unverified = entry({
-      stableKey: 'desktop:ssh:cold',
+      stableKey: 'host:desktop:ssh:cold',
       kind: 'ssh',
       querySupport: 'legacy-unscoped',
       scopeGap: 'target-unverified'
     })
     // Scoped but not yet hydrated: disabled, and no repair to name.
     const unhydrated = entry({
-      stableKey: 'desktop:ssh:warm',
+      stableKey: 'host:desktop:ssh:warm',
       kind: 'ssh',
       owner: null,
       catalogState: 'unhydrated'
@@ -199,19 +199,19 @@ describe('sole create host', () => {
   }
 
   it('states the only eligible host and never one of several', () => {
-    expect(soleAutomationCreateHost([entry()], hydrated)?.stableKey).toBe('desktop:self')
+    expect(soleAutomationCreateHost([entry()], hydrated)?.stableKey).toBe('host:desktop:self')
     expect(soleAutomationCreateHost([entry(), sshEntry(3)], hydrated)).toBeNull()
     // An orphan bucket, an unowned host, and a view-only host are not candidates at all.
     expect(
       soleAutomationCreateHost([entry(), entry({ stableKey: 'orphan', kind: 'orphan' })], hydrated)
         ?.stableKey
-    ).toBe('desktop:self')
+    ).toBe('host:desktop:self')
     expect(
       soleAutomationCreateHost(
         [entry(), entry({ stableKey: 'legacy', querySupport: 'legacy-unscoped' })],
         hydrated
       )?.stableKey
-    ).toBe('desktop:self')
+    ).toBe('host:desktop:self')
   })
 
   it('states nothing until the catalog has settled', () => {
@@ -244,7 +244,7 @@ describe('create project mismatch', () => {
       addedAt: 1,
       worktreeBaseRef: 'main',
       ...overrides
-    } as Repo
+    }
   }
   const desktopSelf = {
     authority: { kind: 'desktop' } as const,

src/renderer/src/components/automations/automation-detail-tab-navigation.test.ts

@@ -68,9 +68,9 @@ describe('shouldHandleAutomationDetailTabArrowKey', () => {
     expect(shouldHandleAutomationDetailTabArrowKey(makeEvent({ target: editable }))).toBe(false)
   })
 
-  it('ignores keys when target is inside a modal dialog, menu, or listbox', () => {
+  it.each(['dialog', 'menu', 'listbox'])('ignores keys inside a %s', (role) => {
     const dialog = document.createElement('div')
-    dialog.setAttribute('role', 'dialog')
+    dialog.setAttribute('role', role)
     const childButton = document.createElement('button')
     dialog.appendChild(childButton)
     document.body.appendChild(dialog)
@@ -198,9 +198,9 @@ describe('shouldHandleAutomationDetailEscapeKey', () => {
     expect(shouldHandleAutomationDetailEscapeKey(makeEvent({ target: clearsValue }))).toBe(false)
   })
 
-  it('ignores Escape when target is inside a modal dialog, menu, or listbox', () => {
+  it.each(['dialog', 'menu', 'listbox'])('ignores Escape inside a %s', (role) => {
     const dialog = document.createElement('div')
-    dialog.setAttribute('role', 'dialog')
+    dialog.setAttribute('role', role)
     const childButton = document.createElement('button')
     dialog.appendChild(childButton)
     document.body.appendChild(dialog)

src/renderer/src/components/automations/automation-extra-agent-args-capability.test.ts

@@ -49,7 +49,7 @@ beforeEach(async () => {
 })
 
 describe('extra agent args capability gate', () => {
-  it('refuses a create or move to a host without the capability before writing', async () => {
+  it('refuses a create on a host without the capability before writing', async () => {
     const { createAutomationForDestination } = await client()
     getRuntimeEnvironmentStatus.mockResolvedValue(LEGACY_HOST)
 

src/renderer/src/components/automations/automation-host-cache-controller.test.ts

@@ -137,7 +137,7 @@ afterEach(() => {
 })
 
 describe('automation host query controller', () => {
-  it('advances catalog generations before fetching anything', async () => {
+  it('syncs catalog generations and fetches the catalog host', async () => {
     const listScoped = vi.fn(() => Promise.resolve(emptyList()))
     const { controller } = createController(listScoped)
     expect(getAuthorityCatalogGeneration(RUNTIME)).toBe(0)

src/renderer/src/components/automations/automation-host-cache.test.ts

@@ -169,16 +169,15 @@ describe('automation host cache invalidation scope', () => {
     expect(cache.freshness(RUNTIME_SELF)).toBe('missing')
   })
 
-  it('evicts a departed entry and caps the retired pool', () => {
+  it('evicts departed entries and rejects work captured before retirement', () => {
     const cache = createCache({ retiredLimit: 1 })
     cache.commit(cache.beginRequest(DESKTOP_SELF), { rows: [row('a')] })
     cache.commit(cache.beginRequest(RUNTIME_SELF), { rows: [row('b')] })
+    const stale = cache.beginRequest(RUNTIME_SELF)
     cache.evict(hostStableKey(DESKTOP_SELF))
     cache.evict(hostStableKey(RUNTIME_SELF))
     expect(cache.keys()).toEqual([])
-    // The revived entry keeps its advanced generation, so nothing captured before eviction commits.
-    const stale = cache.beginRequest(RUNTIME_SELF)
-    cache.invalidate(RUNTIME_SELF)
+    cache.beginRequest(RUNTIME_SELF)
     expect(cache.commit(stale, { rows: [] })).toBe(false)
   })
 

src/renderer/src/components/automations/automation-host-catalog-source.test.ts

@@ -1,4 +1,6 @@
 import { describe, expect, it } from 'vitest'
+import type { RuntimeStatus } from '../../../../shared/runtime-types'
+import { createCompatibleRuntimeStatusResponse } from '@/runtime/runtime-compatibility-test-fixture'
 import { AUTOMATION_LIST_HOST_SCOPE_RUNTIME_CAPABILITY } from '../../../../shared/protocol-version'
 import {
   buildAutomationHostCatalogSource,
@@ -8,6 +10,14 @@ import {
   type AutomationCatalogSshSource
 } from './automation-host-catalog-source'
 
+function status(overrides: Partial<RuntimeStatus> = {}): RuntimeStatus {
+  const response = createCompatibleRuntimeStatusResponse()
+  if (!response.ok) {
+    throw new Error(response.error.message)
+  }
+  return { ...response.result, ...overrides }
+}
+
 function runtime(
   overrides: Partial<AutomationCatalogRuntimeSource> = {}
 ): AutomationCatalogRuntimeSource {
@@ -32,7 +42,7 @@ describe('automation host catalog source', () => {
   it('treats a never-probed environment as loading, not as down', () => {
     expect(runtimeAuthorityHealth(runtime())).toBe('loading')
     expect(runtimeAuthorityHealth(runtime({ status: { status: null } }))).toBe('unavailable')
-    expect(runtimeAuthorityHealth(runtime({ status: { status: {} as never } }))).toBe('fresh')
+    expect(runtimeAuthorityHealth(runtime({ status: { status: status() } }))).toBe('fresh')
   })
 
   it('assumes scoped support until the capability set is actually known', () => {
@@ -40,14 +50,14 @@ describe('automation host catalog source', () => {
     // guessing legacy here would only discard owners on a modern host.
     expect(runtimeQuerySupport(runtime())).toBe('scoped')
     expect(runtimeQuerySupport(runtime({ status: { status: null } }))).toBe('scoped')
-    expect(
-      runtimeQuerySupport(runtime({ status: { status: { capabilities: [] } as never } }))
-    ).toBe('legacy-unscoped')
+    expect(runtimeQuerySupport(runtime({ status: { status: status({ capabilities: [] }) } }))).toBe(
+      'legacy-unscoped'
+    )
     expect(
       runtimeQuerySupport(
         runtime({
           status: {
-            status: { capabilities: [AUTOMATION_LIST_HOST_SCOPE_RUNTIME_CAPABILITY] } as never
+            status: status({ capabilities: [AUTOMATION_LIST_HOST_SCOPE_RUNTIME_CAPABILITY] })
           }
         })
       )

src/renderer/src/components/automations/automation-host-diagnostics.test.ts

@@ -605,16 +605,4 @@ describe('automation host diagnostics bookkeeping', () => {
       byStableKey: {}
     })
   })
-
-  it('never copies an automation, a prompt, or a name into a counter', () => {
-    const diagnostics = createAutomationHostDiagnostics()
-    diagnostics.recordResponse({
-      authorityKey: 'a',
-      stableKey: 'k1',
-      rows: 3,
-      durationMs: 12,
-      approxSerializedChars: 900
-    })
-    expect(JSON.stringify(diagnostics.snapshot())).not.toMatch(/prompt|automation-|secret/i)
-  })
 })

src/renderer/src/components/automations/automation-host-health.test.ts

@@ -45,7 +45,7 @@ describe('execution health', () => {
     expect(resolveSshExecutionHealth('unhydrated', 'connected', 'disconnected')).toBe('unknown')
   })
 
-  it('maps every SSH status onto a distinct execution state', () => {
+  it('maps reported SSH statuses and absent status to execution health', () => {
     expect(resolveSshExecutionHealth('authoritative', 'connected', undefined)).toBe('connected')
     expect(resolveSshExecutionHealth('authoritative', 'reconnecting', undefined)).toBe('connecting')
     expect(resolveSshExecutionHealth('authoritative', 'deploying-relay', undefined)).toBe(

src/renderer/src/components/automations/automation-host-scale-gates.test.ts

@@ -55,8 +55,6 @@ const SCOPED_AUTHORITIES = AUTHORITY_COUNT / 2
 const HOST_COUNT = AUTHORITY_COUNT * ENTRIES_PER_AUTHORITY
 const AUTOMATION_COUNT = HOST_COUNT * AUTOMATIONS_PER_HOST
 const SSH_TARGET_GENERATION = 3
-// Doc 417: a refresh at this size must not block the renderer for a frame budget's worth of work.
-const LONG_TASK_BUDGET_MS = 50
 
 function authorityAt(index: number): AutomationAuthorityRef {
   return { kind: 'runtime', environmentId: `env-${index}`, pairingRevision: index + 1 }
@@ -230,31 +228,6 @@ async function roundTrip<T>(environmentId: string, value: () => T): Promise<T> {
   }
 }
 
-/**
- * Longest gap between successive macrotasks: a synchronous block anywhere in the
- * refresh delays the next timer by its own duration, which is the only way to
- * observe "no long task" rather than to assert it about one function.
- */
-function watchLongTasks(): () => number {
-  let worst = 0
-  let last = performance.now()
-  let running = true
-  const tick = (): void => {
-    if (!running) {
-      return
-    }
-    const at = performance.now()
-    worst = Math.max(worst, at - last)
-    last = at
-    setTimeout(tick, 0)
-  }
-  setTimeout(tick, 0)
-  return () => {
-    running = false
-    return worst
-  }
-}
-
 function harness(): {
   cache: AutomationHostCache
   refresh: (options?: { force?: boolean }) => Promise<void>
@@ -378,15 +351,6 @@ describe('one refresh of 50 hosts carrying 1,000 automations', () => {
     expect(snap.totals.rows).toBe(AUTOMATION_COUNT)
     expect(perHost).toBe(AUTOMATION_COUNT)
   })
-
-  it('does no synchronous work long enough to drop a frame', async () => {
-    const { refresh } = harness()
-    const stop = watchLongTasks()
-
-    await refresh()
-
-    expect(stop()).toBeLessThan(LONG_TASK_BUDGET_MS)
-  })
 })
 
 describe('relay cost the request pool cannot see', () => {
@@ -428,7 +392,7 @@ describe('relay cost the request pool cannot see', () => {
 })
 
 describe('the commit fence under real concurrency', () => {
-  it('commits nothing from an authority that re-paired while its answer was in flight', async () => {
+  it('commits nothing from an authority invalidated while its answer was in flight', async () => {
     const { cache, refresh } = harness()
     // One modern and one old server, so both the per-entry and the fan-out arm
     // are fenced out while the other 40 hosts keep landing around them.

src/renderer/src/components/automations/automation-list-empty-state.test.ts

@@ -209,16 +209,4 @@ describe('automation list empty state', () => {
     expect(state.kind).toBe('host-empty')
     expect(state.title).toBe('No automations on web-01')
   })
-
-  it('keeps each empty string exclusive to its own condition', () => {
-    const titles = [
-      resolveAutomationListEmptyState(input()).title,
-      resolveAutomationListEmptyState(input({ resolution: ALL_HOSTS })).title,
-      resolveAutomationListEmptyState(input({ searchActive: true, hostRowCount: 2 })).title,
-      resolveAutomationListEmptyState(
-        input({ resolution: hostResolution({ entry: entry({ authorityHealth: 'stale-error' }) }) })
-      ).title
-    ]
-    expect(new Set(titles).size).toBe(4)
-  })
 })

src/renderer/src/components/automations/automation-list-search-rows.test.ts

@@ -79,7 +79,9 @@ describe('automation search row fields', () => {
   })
 
   it('indexes external jobs by provider, target host, and workdir', () => {
-    const [entry] = buildExternalAutomationListEntries([makeScopedExternalManager()])
+    const manager = makeScopedExternalManager()
+    manager.manager.jobs[0]!.workdir = '/srv/sweep-project'
+    const [entry] = buildExternalAutomationListEntries([manager])
     if (!entry) {
       throw new Error('fixture produced no external entry')
     }
@@ -88,6 +90,7 @@ describe('automation search row fields', () => {
     expect(fields.agent).toBe('OpenClaw')
     expect(fields.host).toBe('This computer')
     expect(fields.prompt).toBe('Sweep')
+    expect(fields.project).toBe('/srv/sweep-project')
   })
 })
 
@@ -131,14 +134,14 @@ describe('automation search row index', () => {
   })
 
   it('fingerprints identical content identically across a refresh tick', () => {
-    const build = (): ReturnType<typeof buildAutomationSearchRowSources> =>
+    const build = (nextRunAt: number): ReturnType<typeof buildAutomationSearchRowSources> =>
       buildAutomationSearchRowSources(
-        [makeAutomationListRow({ automation: makeAutomation({ nextRunAt: Date.now() }) })],
+        [makeAutomationListRow({ automation: makeAutomation({ nextRunAt }) })],
         { repoMap }
       )
     // Why: nextRunAt churns on every tick and must not invalidate the index.
-    expect(buildAutomationListSearchRowFingerprint(build())).toBe(
-      buildAutomationListSearchRowFingerprint(build())
+    expect(buildAutomationListSearchRowFingerprint(build(100))).toBe(
+      buildAutomationListSearchRowFingerprint(build(200))
     )
   })
 

src/renderer/src/components/automations/automation-run-context.test.ts

@@ -68,7 +68,7 @@ describe('buildAutomationRunContextForRepo', () => {
       buildAutomationRunContextForRepo({
         repoId: 'repo-builder',
         repos: [repo('repo-builder')],
-        projectHostSetups: [setup({ setupState: 'setting-up' })]
+        projectHostSetups: [setup({ hostId: 'local', setupState: 'setting-up' })]
       })
     ).toBeNull()
 

src/shared/command-token-scanner.test.ts

@@ -31,7 +31,7 @@ describe('command token scanner', () => {
     )
   })
 
-  it('extracts path basenames without allocating path segment arrays', () => {
+  it('extracts basenames from Windows and POSIX paths', () => {
     expect(getCommandTokenPathBasename('C:\\Program Files\\Orca\\codex.cmd')).toBe('codex.cmd')
     expect(getCommandTokenPathBasename('/usr/local/bin/omp')).toBe('omp')
   })

src/shared/commit-message-agent-spec.test.ts

@@ -1,4 +1,3 @@
-import { claudeBackgroundArgs } from './claude-background-run-isolation'
 import { afterEach, describe, expect, it, vi } from 'vitest'
 import {
   COMMIT_MESSAGE_AGENT_SPECS,
@@ -26,26 +25,6 @@ afterEach(() => {
 })
 
 describe('COMMIT_MESSAGE_AGENT_SPECS', () => {
-  it('exposes the installed local agents as commit-message agents', () => {
-    const ids = listCommitMessageAgentIds().sort()
-    expect(ids).toEqual([
-      'amp',
-      'antigravity',
-      'claude',
-      'codex',
-      'copilot',
-      'cursor',
-      'dsh',
-      'jcode',
-      'kimi',
-      'muse',
-      'omp',
-      'opencode',
-      'opencode2',
-      'pi'
-    ])
-  })
-
   it('uses the expected defaults for core agents', () => {
     expect(COMMIT_MESSAGE_AGENT_SPECS.claude?.defaultModelId).toBe('sonnet')
     expect(COMMIT_MESSAGE_AGENT_SPECS.codex?.defaultModelId).toBe('gpt-5.6-terra')
@@ -106,28 +85,6 @@ describe('COMMIT_MESSAGE_AGENT_SPECS', () => {
     }
   })
 
-  it('lists Copilot hosted CLI models even when account policy filters the picker', () => {
-    expect(COMMIT_MESSAGE_AGENT_SPECS.copilot?.defaultModelId).toBe('gpt-5.4')
-    expect(COMMIT_MESSAGE_AGENT_SPECS.copilot?.models.map((m) => m.id)).toEqual([
-      'auto',
-      'claude-haiku-4.5',
-      'claude-sonnet-4.5',
-      'claude-sonnet-4.6',
-      'claude-opus-4.5',
-      'claude-opus-4.6',
-      'claude-opus-4.6-fast',
-      'claude-opus-4.7',
-      'gpt-4.1',
-      'gpt-5-mini',
-      'gpt-5.2',
-      'gpt-5.2-codex',
-      'gpt-5.3-codex',
-      'gpt-5.4',
-      'gpt-5.4-mini',
-      'gpt-5.5'
-    ])
-  })
-
   it('treats disabled default agents as unavailable for implicit Source Control AI choices', () => {
     expect(resolveCommitMessageAgentChoice(null, 'codex', ['codex'])).toBe('claude')
     expect(resolveCommitMessageAgentChoice(null, null, ['claude'])).toBeNull()
@@ -174,7 +131,7 @@ describe('COMMIT_MESSAGE_AGENT_SPECS', () => {
     expect(listCommitMessageAgentIds()).not.toContain(CUSTOM_AGENT_ID)
   })
 
-  it('orders Codex models by version descending to match the official picker', () => {
+  it('orders seeded Codex models by version descending', () => {
     const ids = COMMIT_MESSAGE_AGENT_SPECS.codex?.models.map((m) => m.id)
     expect(ids).toEqual([
       'gpt-5.6-terra',
@@ -215,7 +172,9 @@ describe('buildArgs (Claude)', () => {
       'haiku',
       '--permission-mode',
       'plan',
-      ...claudeBackgroundArgs()
+      '--strict-mcp-config',
+      '--settings',
+      '{"disableAllHooks":true}'
     ])
   })
 
@@ -235,7 +194,9 @@ describe('buildArgs (Claude)', () => {
       'plan',
       '--effort',
       'high',
-      ...claudeBackgroundArgs()
+      '--strict-mcp-config',
+      '--settings',
+      '{"disableAllHooks":true}'
     ])
   })
 
@@ -296,7 +257,15 @@ describe('buildArgs (Jcode)', () => {
     const args = spec.buildArgs({ prompt: 'name this branch', model: 'claude-haiku-4-5' })
     const runIndex = args.indexOf('run')
     expect(runIndex).toBeGreaterThan(0)
-    expect(args.slice(0, runIndex).every((arg) => arg.startsWith('--') || arg !== 'run')).toBe(true)
+    expect(args.slice(0, runIndex)).toEqual([
+      '--no-update',
+      '--quiet',
+      '--no-selfdev',
+      '--tool-profile',
+      'none',
+      '--model',
+      'claude-haiku-4-5'
+    ])
     expect(args.slice(runIndex)).toEqual(['run', '--json', 'name this branch'])
   })
 
@@ -385,7 +354,9 @@ describe('model discovery parsers', () => {
       '--output-format',
       'stream-json',
       '--verbose',
-      ...claudeBackgroundArgs()
+      '--strict-mcp-config',
+      '--settings',
+      '{"disableAllHooks":true}'
     ])
     const payload = JSON.parse(discovery?.stdinPayload ?? '') as {
       type?: string

src/shared/commit-message-generation.test.ts

@@ -24,7 +24,7 @@ describe('buildCommitMessagePrompt', () => {
     expect(prompt).not.toContain('Additional user prompt:')
   })
 
-  it('keeps a custom prompt in a separate bounded section', () => {
+  it('keeps a custom prompt in a separate section', () => {
     const prompt = buildCommitMessagePrompt(
       {
         branch: null,

src/shared/commit-message-omp.spec.ts

@@ -18,7 +18,7 @@ describe('OMP Source Control AI', () => {
       })
     }
   )
-  it('delivers large diffs on stdin and keeps generation isolated from tools and extensions', () => {
+  it('plans stdin delivery with tools and extensions disabled', () => {
     const spec = getCommitMessageAgentSpec('omp')
     expect(spec?.promptDelivery).toBe('stdin')
     if (!spec) {

src/shared/conversation-name-generation.test.ts

@@ -6,7 +6,7 @@ import {
 } from './conversation-name-generation'
 
 describe('conversation name generation', () => {
-  it('bounds the first prompt in both the built-in prompt and template variable', () => {
+  it('bounds the first prompt and includes it in the built-in prompt', () => {
     const firstPrompt = `Start ${'x'.repeat(5000)}`
     const clamped = clampConversationNameFirstPrompt(firstPrompt)
     expect(clamped).toHaveLength(4000)

src/shared/darwin-terminal-names.test.ts

@@ -218,10 +218,13 @@ describe('nameDarwinTerminals', () => {
     directory.entries = [
       { name: 'gone', rdev: 0x10000009n, error: new Error('ENOENT') },
       { name: 'denied', rdev: 0x10000009n, error: new Error('EACCES') },
-      { name: 'ttys009', rdev: 0x10000009n }
+      { name: 'cu.found', rdev: 0x10000009n }
     ]
 
-    expect(await nameDarwinTerminals(darwinRow('16/9'), directory)).toBe(darwinRow('ttys009'))
+    expect(await nameDarwinTerminals(darwinRow('16/9'), directory)).toBe(darwinRow('cu.found'))
+    expect(directory.readPaths).toEqual(
+      ['ttys009', 'gone', 'denied', 'cu.found'].map((name) => join('/dev', name))
+    )
   })
 
   it('rejects directory open failures so the reader can fall back to ordinary tty ps', async () => {

src/shared/draft-paste-ready-scanner-grok-trace-replay.test.ts

@@ -62,7 +62,7 @@ function replayReadyAtMs(
 }
 
 describe('grok startup trace replay (alternate-screen mode)', () => {
-  it('delivers on the composer frame instead of waiting out the hard timeout', () => {
+  it('reports readiness within the first second of the alternate-screen trace', () => {
     const readyAt = replayReadyAtMs('grok-composer-prompt', GROK_STARTUP_PTY_TRACE)
     expect(readyAt).not.toBeNull()
     expect(readyAt).toBeLessThan(1000)
@@ -74,7 +74,7 @@ describe('grok startup trace replay (alternate-screen mode)', () => {
     expect(replayReadyAtMs('render-quiet-after-bracketed-paste', GROK_STARTUP_PTY_TRACE)).toBeNull()
   })
 
-  it('fires on the same frame that paints the composer box', () => {
+  it('reports readiness on the frame that paints the composer box', () => {
     const composerFrame = GROK_STARTUP_PTY_TRACE.find((chunk) => chunk.data?.includes('❯'))
     expect(composerFrame).toBeDefined()
     expect(replayReadyAtMs('grok-composer-prompt', GROK_STARTUP_PTY_TRACE)).toBe(composerFrame?.t)
@@ -82,7 +82,7 @@ describe('grok startup trace replay (alternate-screen mode)', () => {
 })
 
 describe('grok startup trace replay (inline mode, no alternate screen)', () => {
-  it('still delivers through the quiet window when the marker never anchors', () => {
+  it('arms a quiet window before the inline trace reaches the modeled timeout', () => {
     // Why: inline grok emits no \x1b[?1049h, so the composer glyph never counts.
     // Anchoring the quiet window on the alt-screen switch too would leave this
     // launch with NO delivery path — orca-runtime drops the draft when readiness
@@ -92,28 +92,9 @@ describe('grok startup trace replay (inline mode, no alternate screen)', () => {
     expect(readyAt).toBeLessThan(HARD_TIMEOUT_MS)
   })
 
-  it('matches the default signal exactly, so inline launches keep their old timing', () => {
+  it('matches the default signal timing for the inline trace', () => {
     expect(replayReadyAtMs('grok-composer-prompt', GROK_INLINE_STARTUP_PTY_TRACE)).toBe(
       replayReadyAtMs('render-quiet-after-bracketed-paste', GROK_INLINE_STARTUP_PTY_TRACE)
     )
   })
-
-  it('reports the hard timeout, not a late marker, once the waiters have settled', () => {
-    // Guards the model above: the real waiters resolve at 8s, so a marker landing
-    // after that must not be reported as a delivery time.
-    const lateMarker: GrokStartupTraceChunk[] = [
-      { t: 0, data: '\x1b[?1049h' },
-      { t: 8500, data: '\x1b[38;2;200;200;200m❯ ' }
-    ]
-    expect(replayReadyAtMs('grok-composer-prompt', lateMarker)).toBeNull()
-  })
-
-  it('records a startup with no alternate-screen switch', () => {
-    // Guards the fixture itself: if a future re-recording captures alt-screen
-    // output, the inline assertions above would silently stop testing inline.
-    expect(GROK_INLINE_STARTUP_PTY_TRACE.some((chunk) => chunk.data?.includes('\x1b[?1049h'))).toBe(
-      false
-    )
-    expect(GROK_INLINE_STARTUP_PTY_TRACE.some((chunk) => chunk.data?.includes('❯'))).toBe(true)
-  })
 })

src/shared/dsh-startup.spec.ts

@@ -4,7 +4,7 @@ import { TUI_AGENT_CONFIG } from './tui-agent-config'
 
 describe('DSH first launch', () => {
   it.each(['darwin', 'linux', 'win32'] as const)(
-    'selects the current workspace before pasting the task on %s',
+    'plans the current workspace with task input deferred on %s',
     (platform) => {
       const plan = buildAgentStartupPlan({
         agent: 'dsh',

src/shared/emoji-shortcode-catalog.lazy.test.ts

@@ -3,8 +3,9 @@ import emojiShortcodes from 'emojibase-data/en/shortcodes/emojibase.json'
 
 async function importConfiguredCatalog() {
   const catalog = await import('./emoji-shortcode-catalog.js')
-  catalog.setEmojiShortcodeDatasetLoader(() => emojiShortcodes)
-  return catalog
+  const loadDataset = vi.fn(() => emojiShortcodes)
+  catalog.setEmojiShortcodeDatasetLoader(loadDataset)
+  return { catalog, loadDataset }
 }
 
 describe('emoji shortcode catalog laziness', () => {
@@ -13,18 +14,21 @@ describe('emoji shortcode catalog laziness', () => {
   })
 
   it('does not build the catalog when the shared module is imported', async () => {
-    const catalog = await importConfiguredCatalog()
+    const { catalog, loadDataset } = await importConfiguredCatalog()
 
-    expect(catalog.isEmojiShortcodeCatalogBuiltForTest()).toBe(false)
+    expect(loadDataset).not.toHaveBeenCalled()
 
-    expect(catalog.getStandardEmojiShortcodeEntries().length).toBeGreaterThan(1000)
-    expect(catalog.isEmojiShortcodeCatalogBuiltForTest()).toBe(true)
+    const entries = catalog.getStandardEmojiShortcodeEntries()
+    expect(entries).toContainEqual({ emoji: '🎉', shortcode: 'party' })
+    expect(catalog.getStandardEmojiShortcodeEntries()).toBe(entries)
+    expect(loadDataset).toHaveBeenCalledTimes(1)
   })
 
-  it('builds on first use and keeps the main process off the eager path', async () => {
-    const catalog = await importConfiguredCatalog()
+  it('loads once when replacing emoji and reuses the catalog', async () => {
+    const { catalog, loadDataset } = await importConfiguredCatalog()
 
     expect(catalog.replaceKnownEmojiWithShortcodes('ship \u{1F389}')).toBe('ship  party ')
-    expect(catalog.isEmojiShortcodeCatalogBuiltForTest()).toBe(true)
+    expect(catalog.replaceKnownEmojiWithShortcodes('again 🎉')).toBe('again  party ')
+    expect(loadDataset).toHaveBeenCalledTimes(1)
   })
 })

src/shared/emulator-keyboard-frame.test.ts

@@ -2,8 +2,7 @@ import { describe, expect, it } from 'vitest'
 import {
   buildServeSimKeyboardFramesForKey,
   buildServeSimKeyboardFramesForText,
-  encodeServeSimKeyboardFrame,
-  SERVE_SIM_KEYBOARD_MESSAGE_TAG
+  encodeServeSimKeyboardFrame
 } from './emulator-keyboard-frame'
 
 describe('serve-sim keyboard frames', () => {
@@ -46,7 +45,7 @@ describe('serve-sim keyboard frames', () => {
   it('encodes serve-sim keyboard messages with the binary keyboard tag', () => {
     const frame = encodeServeSimKeyboardFrame({ type: 'down', usage: 4 })
 
-    expect(frame[0]).toBe(SERVE_SIM_KEYBOARD_MESSAGE_TAG)
+    expect(frame[0]).toBe(0x06)
     expect(JSON.parse(new TextDecoder().decode(frame.subarray(1)))).toEqual({
       type: 'down',
       usage: 4

src/shared/ephemeral-setup-terminal-worktree-id.test.ts

@@ -1,6 +1,5 @@
 import { describe, expect, it } from 'vitest'
 import {
-  EPHEMERAL_SETUP_TERMINAL_WORKTREE_ID_PREFIX,
   brandEphemeralSetupTerminalWorktreeId,
   isEphemeralSetupTerminalWorktreeId
 } from './ephemeral-setup-terminal-worktree-id'
@@ -8,7 +7,7 @@ import {
 describe('ephemeral setup terminal worktree id', () => {
   it('brands a panel id with the ephemeral prefix', () => {
     expect(brandEphemeralSetupTerminalWorktreeId('feature-wall-orchestration-skill-terminal')).toBe(
-      `${EPHEMERAL_SETUP_TERMINAL_WORKTREE_ID_PREFIX}feature-wall-orchestration-skill-terminal`
+      'ephemeral-setup-terminal:feature-wall-orchestration-skill-terminal'
     )
   })