test(web): keyless browser e2e lane — replayed round trip + seeded cold resume

apps/web/tests/harness.ts boots the real web assembly in-process
(startHost llm:false -> installLlmReplay providers-mode -> mountWebPlugins
-> startWebServer) under DSH_SNAPSHOT replay/record/refresh. Barrier
stack: in-process turn/end -> agent.whenIdle (covers the persistence
flush) -> browser settled-poll. Seeding goes through the real persistence
API (semantic-checkpoint precedent); record harvests fixtures from live
session memory and tokenizes {{sessionId}}/{{cwd}}; refresh is the sole
golden writer. Console tripwires fail scenarios on reconnect/gap-repair
self-healing; harness close asserts full replay-fixture consumption.

Scenarios, each with fixtures recorded against THIS assembly via a live
model run: replay-round-trip (real composer -> real bash echo -> settled
markdown + aria golden + world-state event asserts) and seeded-history
(cold sidebar list -> implicit resume on open -> history tool cards from
the log, zero model calls). apps/web/tests are host-plane programs:
excluded from the client-registered apps/web project, included in
tsconfig.host.json (one program cannot hold both Context merge sides).

apps/web/tests/replay-round-trip.e2e.ts

@@ -0,0 +1,109 @@
+// Web e2e scenario: fresh round trip. A real chromium types a prompt into the
+// real composer; the wire, apiproxy, agent loop, and the REAL bash tool (echo
+// in the temp workspace) all run; the model seam is dsh-llm-replay (keyless)
+// or the live adapter (record). Drive steps run in every mode and wait only
+// on generic completion (whenTurnSettled — never model-content selectors, so
+// record cannot hang on a live model answering differently); assertion steps
+// run in replay/refresh only. Settled states only — streaming incrementality
+// is asserted from the persisted assistant/chunk events, not transient DOM.
+// Record: DSH_SNAPSHOT=record rewrites session.jsonl, then a keyless
+// DSH_SNAPSHOT=refresh regenerates ui.expected.md.
+import { readFile } from 'node:fs/promises'
+import { fileURLToPath } from 'node:url'
+import type { Browser, Page } from 'playwright'
+import { chromium } from 'playwright'
+import { afterAll, beforeAll, describe, expect, it, onTestFailed } from 'vitest'
+import type { SessionEvent } from '@deepseek-ai/dsh-session'
+import {
+  assertFixtureInventory, captureStableAria, compareOrRefreshGolden, fixtureUserPrompts,
+  launchWebHarness, recordFixture, watchConsole, webSnapshotMode, type WebHarness,
+} from './harness.ts'
+import { saveFailureShot } from './support.ts'
+
+const SNAPSHOT_DIR = fileURLToPath(new URL('./snapshots/fresh-round-trip', import.meta.url))
+const FIXTURE = fileURLToPath(new URL('./snapshots/fresh-round-trip/session.jsonl', import.meta.url))
+const UI_EXPECTED = fileURLToPath(new URL('./snapshots/fresh-round-trip/ui.expected.md', import.meta.url))
+const MODE = webSnapshotMode()
+
+// The scenario's one drive prompt. Record sends it; replay asserts the
+// committed fixture recorded exactly it, so drive script and fixture cannot
+// drift apart.
+const PROMPT = 'Use the bash tool to run exactly: echo WEB_E2E_OK. Then reply with the single word DONE and stop.'
+
+describe('web e2e: fresh round trip through the real assembly', () => {
+  let harness: WebHarness
+  let browser: Browser
+  let page: Page
+  let tripwire: ReturnType<typeof watchConsole>
+  const sessionEvents: SessionEvent[] = []
+
+  beforeAll(async () => {
+    harness = await launchWebHarness({
+      ...(MODE === 'record' ? {} : { replayFixture: FIXTURE, paceMs: 15 }),
+    })
+    harness.host.ctx.on('session/event', (_session, event: SessionEvent) => { sessionEvents.push(event) })
+    browser = await chromium.launch()
+    page = await browser.newPage({ viewport: { width: 1680, height: 1000 } })
+    tripwire = watchConsole(page)
+    await page.goto(harness.baseUrl, { waitUntil: 'load' })
+    await page.waitForSelector('[class*="frame"]', { timeout: 30_000 })
+  }, 120_000)
+
+  afterAll(async () => {
+    await browser?.close()
+    await harness?.close()
+  })
+
+  it('drives the recorded prompt to a settled turn (all modes)', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-round-trip'))
+    if (MODE !== 'record') {
+      // Drift guard: the committed fixture must carry exactly the drive prompt.
+      expect(fixtureUserPrompts(await readFile(FIXTURE, 'utf8'))).toEqual([PROMPT])
+    }
+    const input = page.locator('textarea').first()
+    await input.waitFor({ timeout: 10_000 })
+    // Arm the host-side settled barrier BEFORE the send click.
+    const settled = harness.whenTurnSettled()
+    await input.fill(PROMPT)
+    await input.press('Enter')
+    const sessionId = await settled
+    if (MODE === 'record') {
+      await recordFixture(harness, sessionId, FIXTURE)
+    }
+  }, 200_000)
+
+  it.skipIf(MODE === 'record')('rendered the settled turn: markdown, tool row, composer restore', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-round-trip-settled'))
+    // Browser settled-poll after host completion (host strictly precedes render).
+    await page.locator('[data-streaming="true"]').waitFor({ state: 'detached', timeout: 15_000 }).catch(() => {
+      // Chunks may coalesce into one commit; a never-mounted streaming node is
+      // legal — the chunk-event assertions below carry incrementality.
+    })
+    await expect.poll(() => page.getByText('DONE', { exact: true }).count(), { timeout: 15_000 }).toBeGreaterThanOrEqual(1)
+    // World state, not self-report: bash really ran and the turn closed clean.
+    const toolCalls = sessionEvents.filter(e => e.type === 'tool/call')
+    expect(toolCalls.map(e => (e as SessionEvent & { data: { name: string } }).data.name)).toContain('bash')
+    const turnEnds = sessionEvents.filter(e => e.type === 'turn/end')
+    expect(turnEnds.length).toBe(1)
+    expect((turnEnds[0] as SessionEvent & { data: { reason: { kind: string } } }).data.reason.kind).toBe('completed')
+    // The persisted chunk events are the authoritative incrementality proof.
+    expect(sessionEvents.filter(e => e.type === 'assistant/chunk').length).toBeGreaterThan(10)
+  }, 60_000)
+
+  it.skipIf(MODE === 'record')('matches the conversation aria golden with stable anchors', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-round-trip-aria'))
+    // Anchor assertions survive a semantics-preserving component rewrite even
+    // while the whole-region golden churns.
+    await expect(page.getByRole('textbox').first().isVisible()).resolves.toBe(true)
+    expect(await page.getByText('WEB_E2E_OK', { exact: false }).count()).toBeGreaterThanOrEqual(1)
+    const snapshot = await captureStableAria(page, '[class*="centerCol"]', harness.workspaceCwd)
+    await compareOrRefreshGolden(UI_EXPECTED, snapshot, MODE)
+  })
+
+  it.skipIf(MODE === 'record')('stayed clean: no pageerrors, no reconnect self-healing, no server errors', async () => {
+    expect(tripwire.pageErrors).toEqual([])
+    expect(tripwire.warnings).toEqual([])
+    expect(harness.serverErrors).toEqual([])
+    await assertFixtureInventory(SNAPSHOT_DIR, ['session.jsonl', 'ui.expected.md'])
+  })
+})