Desktop release (umbrella): meetings STT + capture grounding + remaining todos

e2e/meeting-transcription.spec.ts

@@ -0,0 +1,65 @@
+/**
+ * E2E: the Meetings detail exposes transcription provenance (which STT engine + model ran)
+ * against the real built app + seeded synthetic meetings. The Re-transcribe control and its
+ * failure feedback are behavior-verified in the MeetingsScreen integration test (they require a
+ * saved recording; seeded meetings are transcript-only), so here we verify the provenance the
+ * user sees and capture a screenshot for the record.
+ */
+import { test, expect, _electron as electron, type ElectronApplication, type Page } from '@playwright/test'
+import os from 'os'
+import path from 'path'
+import fs from 'fs'
+
+const PRO_PRESENT = fs.existsSync(path.resolve('pro/package.json'))
+let app: ElectronApplication
+let page: Page
+let userDataDir: string
+
+test.beforeAll(async () => {
+  test.skip(!PRO_PRESENT, 'pro package not present')
+  userDataDir = fs.mkdtempSync(path.join(os.tmpdir(), 'offgrid-meet-'))
+  app = await electron.launch({
+    args: ['.'],
+    env: {
+      ...process.env,
+      OFFGRID_USER_DATA: userDataDir,
+      OFFGRID_PRO: '1',
+      OFFGRID_SEED: 'force',
+      OFFGRID_SEED_PRO: 'force',
+      NODE_ENV: 'production'
+    }
+  })
+  page = await app.firstWindow()
+  await page.emulateMedia({ reducedMotion: 'reduce' })
+  await page.waitForLoadState('domcontentloaded')
+  for (let i = 0; i < 8; i++) {
+    const btn = page.getByRole('button', { name: /Continue|Start using Off Grid/i })
+    if (!(await btn.isVisible().catch(() => false))) break
+    await btn.click().catch(() => {})
+    await page.waitForTimeout(400)
+  }
+  await page.waitForTimeout(1500)
+})
+
+test.afterAll(async () => {
+  await app?.close()
+  fs.rmSync(userDataDir, { recursive: true, force: true })
+})
+
+test('meeting detail exposes the STT model picker (view + change)', async () => {
+  await page.getByRole('button', { name: 'Meetings', exact: true }).first().click()
+  await page.waitForTimeout(800)
+  // Open the first seeded meeting so the detail (with the transcription controls) renders.
+  const firstMeeting = page.locator('[class*="cursor-pointer"]').filter({ hasText: /·/ }).first()
+  if (await firstMeeting.isVisible().catch(() => false)) {
+    await firstMeeting.click().catch(() => {})
+  }
+  await page.waitForTimeout(600)
+  // The model picker is read from the core transcription source of truth and lets the user see
+  // (and change) which STT model transcribes meetings — always visible on the detail.
+  const picker = page.getByLabel('Transcription model')
+  await expect(picker).toBeVisible({ timeout: 8000 })
+  // At minimum the built-in whisper default is offered (installed models add more in a real profile).
+  await expect(picker).toContainText(/Whisper/)
+  await page.screenshot({ path: 'e2e/screenshots/meeting-transcription-picker.png' })
+})