feat(app): v2 review panel overhaul (#31882)

Co-authored-by: LukeParkerDev <10430890+Hona@users.noreply.github.com>
2026-07-02 13:11:58 +05:30
parent 0b5420391e
commit de915b7205
35 changed files with 3438 additions and 214 deletions
@@ -2,7 +2,13 @@ import type { Page } from "@playwright/test"
 import { expectSessionTitle } from "../../utils/waits"
 import { benchmark, expect, withBenchmarkPage } from "../benchmark"
 import { fixture } from "./session-timeline-stress.fixture"
-import { installStressSessionTabs, mockStressTimeline, stressSessionHref } from "./timeline-test-helpers"
+import {
+  createReviewDiffs,
+  installStressSessionTabs,
+  installTimelineSettings,
+  mockStressTimeline,
+  stressSessionHref,
+} from "./timeline-test-helpers"
 import { measureSessionSwitch, waitForStableTimeline } from "./session-tab-switch-probe"

 type Result = Awaited<ReturnType<typeof measureSessionSwitch>>
@@ -20,8 +26,41 @@ benchmark("benchmarks cold and hot session tab switching", async ({ browser, rep
  report({ results, summary: summarize(results) })
 })

-async function trial(page: Page, mode: "cold" | "hot") {
-  await mockStressTimeline(page)
+benchmark(
+  "benchmarks v2 session tab switching with and without the review pane",
+  async ({ browser, report }, testInfo) => {
+    benchmark.setTimeout(360_000)
+    const runs = Number(process.env.SESSION_TAB_SWITCH_RUNS ?? 5)
+    const results = {
+      closed: { cold: [] as Result[], hot: [] as Result[] },
+      open: { cold: [] as Result[], hot: [] as Result[] },
+    }
+    for (const reviewPane of ["closed", "open"] as const) {
+      for (const mode of ["cold", "hot"] as const) {
+        for (let run = 0; run < runs; run++) {
+          results[reviewPane][mode].push(
+            await withBenchmarkPage(
+              browser,
+              `session-tab-switch-v2-${reviewPane}-${mode}-${run}`,
+              (page) => trial(page, mode, { newLayoutDesigns: true, reviewPane }),
+              testInfo,
+            ),
+          )
+        }
+      }
+    }
+    report({ results, summary: summarizeReviewPane(results) }, { runs, reviewDiffs: createReviewDiffs().length })
+  },
+)
+
+async function trial(
+  page: Page,
+  mode: "cold" | "hot",
+  options?: { newLayoutDesigns?: boolean; reviewPane?: "closed" | "open" },
+) {
+  const reviewDiffs = options?.newLayoutDesigns ? createReviewDiffs() : undefined
+  await mockStressTimeline(page, { vcsDiff: reviewDiffs })
+  if (options?.newLayoutDesigns) await installTimelineSettings(page)
  await installStressSessionTabs(page)
  if (mode === "hot") {
    await page.goto(stressSessionHref(fixture.targetID))
@@ -33,6 +72,10 @@ async function trial(page: Page, mode: "cold" | "hot") {
    await expectSessionTitle(page, fixture.expected.sourceTitle)
  }
  await waitForStableTimeline(page, fixture.expected.sourceMessageIDs.at(-1)!)
+  if (options?.reviewPane === "open") {
+    await openReviewPane(page)
+    await waitForStableTimeline(page, fixture.expected.sourceMessageIDs.at(-1)!)
+  }

  const destinationIDs = fixture.messages[fixture.targetID].map((message) => message.info.id)
  const sourceIDs = fixture.messages[fixture.sourceID].map((message) => message.info.id)
@@ -70,6 +113,15 @@ function summarize(results: Record<"cold" | "hot", Result[]>) {
  )
 }

+function summarizeReviewPane(results: Record<"closed" | "open", Record<"cold" | "hot", Result[]>>) {
+  return Object.fromEntries(
+    Object.entries(results).map(([reviewPane, values]) => [
+      reviewPane,
+      summarize(values as Record<"cold" | "hot", Result[]>),
+    ]),
+  )
+}
+
 async function switchSession(page: Page, sessionID: string, title: string) {
  const href = stressSessionHref(sessionID)
  const tab = page.locator(`[data-slot="titlebar-tabs"] a[href="${href}"]`).first()
@@ -77,3 +129,16 @@ async function switchSession(page: Page, sessionID: string, title: string) {
  await tab.click()
  await expectSessionTitle(page, title)
 }
+
+async function openReviewPane(page: Page) {
+  await page.getByRole("button", { name: "Toggle review" }).click()
+  const panel = page.locator("#review-panel")
+  await expect(panel).toBeVisible()
+  // Text-based readiness works across review implementations; the legacy list mounts
+  // diff viewers lazily while V2 mounts the active preview eagerly.
+  await page.waitForFunction(() => {
+    const panel = document.querySelector<HTMLElement>("#review-panel")
+    const text = panel?.textContent ?? ""
+    return text.includes("generated-000.ts") && text.includes("+3")
+  })
+}
@@ -5,6 +5,12 @@ export type SessionSwitchSample = {
  hasVisibleRows: boolean
  last: boolean
  bottomErrorPx?: number
+  review?: {
+    fileHost: boolean
+    fileHostReplaced: boolean
+    header: string
+    replacedLevels: string[]
+  }
 }

 export function classifySessionSwitch(samples: SessionSwitchSample[]) {
@@ -23,6 +29,10 @@ export function classifySessionSwitch(samples: SessionSwitchSample[]) {
      (sample) => sample.hasVisibleRows && sample.destination.length === 0 && sample.source.length === 0,
    ).length,
    sourceSamples: samples.filter((sample) => sample.source.length > 0).length,
+    reviewFileHostMissingSamples: samples.filter((sample) => sample.review && !sample.review.fileHost).length,
+    reviewFileHostReplacedSamples: samples.filter((sample) => sample.review?.fileHostReplaced).length,
+    reviewHeaders: [...new Set(samples.flatMap((sample) => (sample.review ? [sample.review.header] : [])))],
+    reviewReplacedLevels: [...new Set(samples.flatMap((sample) => sample.review?.replacedLevels ?? []))],
  }
 }

@@ -16,11 +16,41 @@ async function installSessionSwitchProbe(
    const samples: SessionSwitchSample[] = []
    let started: number | undefined
    let running = true
+    const reviewLevels: Record<string, string> = {
+      panel: "#review-panel",
+      tabs: '#review-panel [data-component="tabs"]',
+      body: '#review-panel [data-slot="session-review-v2-body"]',
+      review: '#review-panel [data-component="session-review-v2"]',
+      preview: '#review-panel [data-slot="session-review-v2-preview"]',
+      scroll: '#review-panel [data-slot="session-review-v2-diff-scroll"]',
+      file: '#review-panel [data-component="file"][data-mode="diff"]',
+    }
+    const initialReviewNodes: Record<string, Element | null> = {}
    const sample = () => {
      if (!running || started === undefined) return
      setTimeout(() => {
        if (!running || started === undefined) return
        const observedAtMs = performance.now() - started
+        const reviewPanel = document.querySelector<HTMLElement>("#review-panel")
+        const reviewFile = reviewPanel?.querySelector('[data-component="file"][data-mode="diff"]')
+        const initialReviewFile = initialReviewNodes.file
+        const replacedLevels = Object.entries(reviewLevels).flatMap(([name, selector]) => {
+          const initial = initialReviewNodes[name]
+          if (!initial) return []
+          const current = document.querySelector(selector)
+          return current && current !== initial ? [name] : []
+        })
+        const review = reviewPanel
+          ? {
+              fileHost: !!reviewFile,
+              fileHostReplaced: !!initialReviewFile && !!reviewFile && reviewFile !== initialReviewFile,
+              header:
+                reviewPanel
+                  .querySelector<HTMLElement>('[data-slot="session-review-v2-file-header"]')
+                  ?.textContent?.trim() ?? "",
+              replacedLevels,
+            }
+          : undefined
        const root = [...document.querySelectorAll<HTMLElement>(".scroll-view__viewport")].find((element) =>
          element.querySelector("[data-timeline-row]"),
        )
@@ -44,9 +74,10 @@ async function installSessionSwitchProbe(
            hasVisibleRows,
            last: visible.includes(lastID),
            bottomErrorPx: spacer ? spacer.bottom - view.bottom : undefined,
+            review,
          })
        } else {
-          samples.push({ observedAtMs, destination: [], source: [], hasVisibleRows: false, last: false })
+          samples.push({ observedAtMs, destination: [], source: [], hasVisibleRows: false, last: false, review })
        }
        requestAnimationFrame(sample)
      }, 0)
@@ -57,6 +88,9 @@ async function installSessionSwitchProbe(
        const link = event.target instanceof Element ? event.target.closest("a") : undefined
        if (link?.getAttribute("href") !== href) return
        started = performance.now()
+        for (const [name, selector] of Object.entries(reviewLevels)) {
+          initialReviewNodes[name] = document.querySelector(selector)
+        }
        requestAnimationFrame(sample)
      },
      { capture: true, once: true },
@@ -93,36 +93,53 @@ const assistantMessage = {
  parts: [editPart],
 }

-export async function setupTimelineBenchmark(page: Page, options: { historyTurns: number; eventBatch: number }) {
+export async function setupTimelineBenchmark(
+  page: Page,
+  options: {
+    historyTurns: number
+    eventBatch: number
+    newLayoutDesigns?: boolean
+    vcsDiff?: unknown[]
+    turnDiffs?: unknown[]
+  },
+) {
  const events: EventPayload[] = []
  let eventBatch = options.eventBatch
+  const currentUserMessage = options.turnDiffs
+    ? { ...userMessage, info: { ...userMessage.info, summary: { diffs: options.turnDiffs } } }
+    : userMessage
  await mockOpenCodeServer(page, {
    directory,
    project: project(),
    provider: provider(),
    sessions: [session()],
+    vcsDiff: options.vcsDiff,
    pageMessages: () => ({
      items: [
        ...Array.from({ length: options.historyTurns }, (_, index) => performanceTurn(index)).flat(),
-        userMessage,
+        currentUserMessage,
        assistantMessage,
      ],
    }),
    events: () => events.splice(0, eventBatch),
    eventRetry: 16,
  })
-  await page.addInitScript(() => {
-    localStorage.setItem(
-      "settings.v3",
-      JSON.stringify({
-        general: {
-          editToolPartsExpanded: true,
-          shellToolPartsExpanded: true,
-          showReasoningSummaries: true,
-        },
-      }),
-    )
-  })
+  await page.addInitScript(
+    (input) => {
+      localStorage.setItem(
+        "settings.v3",
+        JSON.stringify({
+          general: {
+            newLayoutDesigns: input.newLayoutDesigns,
+            editToolPartsExpanded: true,
+            shellToolPartsExpanded: true,
+            showReasoningSummaries: true,
+          },
+        }),
+      )
+    },
+    { newLayoutDesigns: options.newLayoutDesigns ?? false },
+  )
  await page.setViewportSize({ width: 1366, height: 768 })
  const scroller = page.locator(".scroll-view__viewport", { has: page.locator("[data-timeline-row]") })
  const text = page.locator(`[data-timeline-part-id="${textPartID}"]`).first()
@@ -1,3 +1,4 @@
+import type { Page } from "@playwright/test"
 import { benchmark, benchmarkDiagnostics, expect } from "../benchmark"
 import {
  buildInitialStreamEvent,
@@ -6,80 +7,300 @@ import {
  textPartID,
 } from "./session-timeline-benchmark.fixture"
 import { startTimelineProfile } from "./session-timeline-profile"
+import { createReviewDiffs } from "./timeline-test-helpers"
 import {
  collectTimelineStreamMetrics,
  installTimelineStreamProbe,
  startTimelineStreamProbe,
 } from "./session-timeline-stream-probe"

+type TimelineStreamOptions = {
+  newLayoutDesigns?: boolean
+  reviewDiffs?: boolean
+  reviewPane?: boolean
+}
+
+type ReviewPaneSample = {
+  observedAtMs: number
+  panelVisible: boolean
+  header: string
+  diffViewers: number
+  diffLines: number
+  codeBlocks: number
+  ready: boolean
+}
+
+type ReviewPaneProbe = {
+  samples: ReviewPaneSample[]
+  start: () => void
+  stop: () => void
+}
+
+const reviewReadyStreak = 3
+
 benchmark.describe("performance: session timeline streaming", () => {
  benchmark("streams assistant text without remounting or oscillating", async ({ page, report }) => {
-    benchmark.setTimeout(480_000)
-    const cpuThrottle = Number(process.env.TIMELINE_CPU_THROTTLE ?? 30)
-    const deltaCount = Number(process.env.TIMELINE_DELTA_COUNT ?? 160)
-    const historyTurns = Number(process.env.TIMELINE_HISTORY_TURNS ?? 320)
-    const eventBatch = Number(process.env.TIMELINE_EVENT_BATCH ?? 1)
-    const minimal = process.env.TIMELINE_MINIMAL === "1"
-    const profileCPU = process.env.TIMELINE_CPU_PROFILE === "1"
-    const profileVisual = !minimal && profileCPU && process.env.TIMELINE_VISUAL_PROFILE !== "0"
+    benchmark.setTimeout(Number(process.env.TIMELINE_COMPLETION_TIMEOUT_MS ?? 420_000) + 60_000)
+    const result = await runTimelineStreamBenchmark(page, {})
+    report(result.metrics, result.context)
+  })
+
+  benchmark("streams assistant text in v2 with review pane closed", async ({ page, report }) => {
+    benchmark.setTimeout(Number(process.env.TIMELINE_COMPLETION_TIMEOUT_MS ?? 420_000) + 60_000)
+    const result = await runTimelineStreamBenchmark(page, { newLayoutDesigns: true })
+    report(result.metrics, result.context)
+  })
+
+  benchmark("streams assistant text in v2 with review diffs and pane closed", async ({ page, report }) => {
+    benchmark.setTimeout(Number(process.env.TIMELINE_COMPLETION_TIMEOUT_MS ?? 420_000) + 60_000)
+    const result = await runTimelineStreamBenchmark(page, { newLayoutDesigns: true, reviewDiffs: true })
+    report(result.metrics, result.context)
+  })
+
+  benchmark("streams assistant text in v2 with review pane open", async ({ page, report }) => {
+    benchmark.setTimeout(Number(process.env.TIMELINE_COMPLETION_TIMEOUT_MS ?? 420_000) + 60_000)
+    const result = await runTimelineStreamBenchmark(page, { newLayoutDesigns: true, reviewPane: true })
+    report(result.metrics, result.context)
+  })
+})
+
+benchmark.describe("performance: review pane", () => {
+  benchmark("loads v2 review diffs and switches active files", async ({ page, report }) => {
+    benchmark.setTimeout(240_000)
+    const historyTurns = Number(process.env.REVIEW_PANE_HISTORY_TURNS ?? 72)
+    const diffs = createReviewDiffs()
    const fixture = await setupTimelineBenchmark(page, {
      historyTurns,
-      eventBatch,
+      eventBatch: 1,
+      newLayoutDesigns: true,
+      vcsDiff: diffs,
    })

-    fixture.transport.enqueue(buildInitialStreamEvent(deltaCount))
-    const contentStart = performance.now()
+    fixture.transport.enqueue(buildInitialStreamEvent(1))
    await expect(fixture.text).toBeVisible()
    await expect(fixture.text).toContainText("Implementation plan")
-    const initialContentObservedMs = performance.now() - contentStart
    await fixture.scrollToBottom()
    await fixture.waitForStableGeometry()

-    const profile = await startTimelineProfile(page, { cpuThrottle, profileCPU })
-    await installTimelineStreamProbe(page, { textPartID, finalIndex: deltaCount, profileVisual, minimal })
-    const deltas = buildStreamDeltaEvents(deltaCount)
-    await startTimelineStreamProbe(page)
-    fixture.transport.enqueue(deltas)
-
-    await page.waitForFunction(
-      (finalIndex) =>
-        (
-          window as Window & {
-            __timelineStreamBenchmark?: { applied: { index: number }[] }
-          }
-        ).__timelineStreamBenchmark?.applied.some((value) => value.index === finalIndex),
-      deltaCount,
-      { timeout: 420_000 },
-    )
-    await expect(fixture.text).toContainText("benchmark-complete")
-    await expect(fixture.text).toContainText("Streaming")
-    await fixture.waitForStableGeometry()
-    const metrics = await collectTimelineStreamMetrics(page, {
-      textPartID,
-      finalIndex: deltaCount,
-      navigations: benchmarkDiagnostics(page).navigations,
-    })
-    const delivered = deltas.length - fixture.transport.pendingCount()
-    await profile.stop()
+    const open = await measureReviewPaneLoad(page, diffs[0]!.file)
+    const switches = []
+    for (const diff of diffs.slice(1, 4)) switches.push(await measureReviewNextFile(page, diff.file))

    report(
      {
-        endToEndInitialContentObservedMs: initialContentObservedMs,
-        ...metrics,
-        deliveredDeltas: delivered,
-        pendingDeltas: fixture.transport.pendingCount(),
+        open,
+        switches,
      },
      {
-        cpuThrottle,
-        profileCPU,
-        profileVisual,
-        minimal,
-        queuedDeltas: deltas.length,
        historyTurns,
-        eventBatch,
+        reviewDiffs: diffs.length,
      },
    )
-
-    await profile.reset()
  })
 })
+
+async function runTimelineStreamBenchmark(page: Page, options: TimelineStreamOptions) {
+  const completionTimeoutMs = Number(process.env.TIMELINE_COMPLETION_TIMEOUT_MS ?? 420_000)
+  const cpuThrottle = Number(process.env.TIMELINE_CPU_THROTTLE ?? 30)
+  const deltaCount = Number(process.env.TIMELINE_DELTA_COUNT ?? 160)
+  const historyTurns = Number(process.env.TIMELINE_HISTORY_TURNS ?? 320)
+  const eventBatch = Number(process.env.TIMELINE_EVENT_BATCH ?? 1)
+  const minimal = process.env.TIMELINE_MINIMAL === "1"
+  const profileCPU = process.env.TIMELINE_CPU_PROFILE === "1"
+  const profileVisual = !minimal && profileCPU && process.env.TIMELINE_VISUAL_PROFILE !== "0"
+  const diffs = options.reviewDiffs || options.reviewPane ? createReviewDiffs() : undefined
+  const fixture = await setupTimelineBenchmark(page, {
+    historyTurns,
+    eventBatch,
+    newLayoutDesigns: options.newLayoutDesigns,
+    // Turn diffs exercise timeline data cost; the pane-open scenario serves the same
+    // diffs through the default git mode so it works across review implementations.
+    turnDiffs: options.reviewDiffs ? diffs : undefined,
+    vcsDiff: options.reviewPane ? diffs : undefined,
+  })
+
+  fixture.transport.enqueue(buildInitialStreamEvent(deltaCount))
+  const contentStart = performance.now()
+  await expect(fixture.text).toBeVisible()
+  await expect(fixture.text).toContainText("Implementation plan")
+  const initialContentObservedMs = performance.now() - contentStart
+  await fixture.scrollToBottom()
+  await fixture.waitForStableGeometry()
+
+  const reviewPane = options.reviewPane && diffs ? await measureReviewPaneLoad(page, diffs[0]!.file) : undefined
+  if (reviewPane) await fixture.waitForStableGeometry()
+
+  const profile = await startTimelineProfile(page, { cpuThrottle, profileCPU })
+  await installTimelineStreamProbe(page, { textPartID, finalIndex: deltaCount, profileVisual, minimal })
+  const deltas = buildStreamDeltaEvents(deltaCount)
+  await startTimelineStreamProbe(page)
+  fixture.transport.enqueue(deltas)
+
+  await page.waitForFunction(
+    (finalIndex) =>
+      (
+        window as Window & {
+          __timelineStreamBenchmark?: { applied: { index: number }[] }
+        }
+      ).__timelineStreamBenchmark?.applied.some((value) => value.index === finalIndex),
+    deltaCount,
+    { timeout: completionTimeoutMs },
+  )
+  await expect(fixture.text).toContainText("benchmark-complete")
+  await expect(fixture.text).toContainText("Streaming")
+  await fixture.waitForStableGeometry()
+  const metrics = await collectTimelineStreamMetrics(page, {
+    textPartID,
+    finalIndex: deltaCount,
+    navigations: benchmarkDiagnostics(page).navigations,
+  })
+  const delivered = deltas.length - fixture.transport.pendingCount()
+  await profile.stop()
+
+  const result = {
+    metrics: {
+      endToEndInitialContentObservedMs: initialContentObservedMs,
+      ...metrics,
+      deliveredDeltas: delivered,
+      pendingDeltas: fixture.transport.pendingCount(),
+      reviewPane: reviewPane ?? null,
+    },
+    context: {
+      cpuThrottle,
+      profileCPU,
+      profileVisual,
+      minimal,
+      queuedDeltas: deltas.length,
+      historyTurns,
+      eventBatch,
+      newLayoutDesigns: options.newLayoutDesigns === true,
+      reviewPane: options.reviewPane === true ? "open" : "closed",
+      reviewDiffs: diffs?.length ?? 0,
+    },
+  }
+
+  await profile.reset()
+  return result
+}
+
+async function measureReviewPaneLoad(page: Page, file: string) {
+  // Default git mode reads the mocked /vcs/diff data, so opening the pane is enough
+  // and the flow works across review pane implementations.
+  await installReviewPaneProbe(page, { file })
+  await startReviewPaneProbe(page)
+  await page.getByRole("button", { name: "Toggle review" }).click()
+  await expect(page.locator("#review-panel")).toBeVisible()
+  return collectReviewPaneProbe(page)
+}
+
+async function measureReviewNextFile(page: Page, file: string) {
+  await installReviewPaneProbe(page, { file })
+  await startReviewPaneProbe(page)
+  await page.getByRole("button", { name: "Next file" }).click()
+  return collectReviewPaneProbe(page)
+}
+
+async function installReviewPaneProbe(page: Page, input: { file: string }) {
+  await page.evaluate((input) => {
+    const samples: ReviewPaneSample[] = []
+    const basename = input.file.split(/[\\/]/).at(-1) ?? input.file
+    let started: number | undefined
+    let running = true
+
+    const paneState = () => {
+      const panel = document.querySelector<HTMLElement>("#review-panel")
+      const review = panel?.querySelector<HTMLElement>('[data-component="session-review-v2"]')
+      const rect = (review ?? panel)?.getBoundingClientRect()
+      const text = panel?.textContent ?? ""
+      const previewHeader = panel?.querySelector<HTMLElement>(
+        '[data-slot="session-review-v2-file-header"]',
+      )?.textContent
+      const header = previewHeader ?? text
+      const viewers = panel ? [...panel.querySelectorAll<HTMLElement>('[data-component="file"][data-mode="diff"]')] : []
+      const codeBlocks = panel?.querySelectorAll("code").length ?? 0
+      const diffLines = viewers.reduce(
+        (sum, viewer) =>
+          sum +
+          (viewer.shadowRoot?.querySelectorAll("[data-line]").length ?? viewer.querySelectorAll("[data-line]").length),
+        0,
+      )
+      const panelVisible =
+        !!panel && panel.getAttribute("aria-hidden") !== "true" && !!rect && rect.width > 0 && rect.height > 0
+      return {
+        panelVisible,
+        header: header.slice(0, 500),
+        diffViewers: viewers.length,
+        diffLines,
+        codeBlocks,
+        ready:
+          panelVisible &&
+          header.includes(basename) &&
+          (viewers.length > 0 || text.includes("+3") || diffLines > 0 || codeBlocks > 0),
+      }
+    }
+
+    const sample = () => {
+      if (!running || started === undefined) return
+      requestAnimationFrame(() => {
+        setTimeout(() => {
+          if (!running || started === undefined) return
+          samples.push({ observedAtMs: performance.now() - started, ...paneState() })
+          if (performance.now() - started < 10_000) sample()
+        }, 0)
+      })
+    }
+
+    ;(window as Window & { __reviewPaneProbe?: ReviewPaneProbe }).__reviewPaneProbe = {
+      samples,
+      start: () => {
+        started = performance.now()
+        performance.mark("opencode.review-pane.click")
+        sample()
+      },
+      stop: () => {
+        running = false
+      },
+    }
+  }, input)
+}
+
+async function startReviewPaneProbe(page: Page) {
+  await page.evaluate(() => {
+    ;(window as Window & { __reviewPaneProbe?: ReviewPaneProbe }).__reviewPaneProbe!.start()
+  })
+}
+
+async function collectReviewPaneProbe(page: Page) {
+  await page.waitForFunction((streak) => {
+    const samples = (window as Window & { __reviewPaneProbe?: ReviewPaneProbe }).__reviewPaneProbe?.samples
+    if (!samples) return false
+    return samples.some((_, index) => {
+      const stable = samples.slice(index, index + streak)
+      return stable.length === streak && stable.every((sample) => sample.ready)
+    })
+  }, reviewReadyStreak)
+
+  const samples = await page.evaluate(() => {
+    const probe = (window as Window & { __reviewPaneProbe?: ReviewPaneProbe }).__reviewPaneProbe!
+    probe.stop()
+    return probe.samples
+  })
+  return { summary: summarizeReviewPaneSamples(samples), samples }
+}
+
+function summarizeReviewPaneSamples(samples: ReviewPaneSample[]) {
+  const firstReady = samples.find((sample) => sample.ready)
+  const stableIndex = samples.findIndex((_, index) => {
+    const stable = samples.slice(index, index + reviewReadyStreak)
+    return stable.length === reviewReadyStreak && stable.every((sample) => sample.ready)
+  })
+  return {
+    samples: samples.length,
+    firstReadyObservedMs: firstReady?.observedAtMs ?? null,
+    stableReadyObservedMs: stableIndex === -1 ? null : samples[stableIndex + reviewReadyStreak - 1]!.observedAtMs,
+    notReadySamples: samples.filter((sample) => !sample.ready).length,
+    maxDiffViewers: Math.max(0, ...samples.map((sample) => sample.diffViewers)),
+    maxDiffLines: Math.max(0, ...samples.map((sample) => sample.diffLines)),
+    maxCodeBlocks: Math.max(0, ...samples.map((sample) => sample.codeBlocks)),
+  }
+}
@@ -21,7 +21,10 @@ export async function installTimelineSettings(page: Page) {

 export function mockStressTimeline(
  page: Page,
-  input?: { onMessages?: (input: { sessionID: string; before?: string; phase: "start" | "end" }) => void },
+  input?: {
+    onMessages?: (input: { sessionID: string; before?: string; phase: "start" | "end" }) => void
+    vcsDiff?: unknown[]
+  },
 ) {
  return mockOpenCodeServer(page, {
    sessions: fixture.sessions,
@@ -30,6 +33,7 @@ export function mockStressTimeline(
    project: fixture.project,
    pageMessages,
    onMessages: input?.onMessages,
+    vcsDiff: input?.vcsDiff,
  })
 }

@@ -78,3 +82,53 @@ export function stressDraftHref(draftID: string) {
 function stressServer() {
  return `http://${process.env.PLAYWRIGHT_SERVER_HOST ?? "127.0.0.1"}:${process.env.PLAYWRIGHT_SERVER_PORT ?? "4096"}`
 }
+
+export function createReviewDiffs() {
+  return Array.from({ length: Number(process.env.REVIEW_PANE_DIFF_COUNT ?? 72) }, (_, index) => {
+    const lines = index % 3 === 0 ? 300 : index % 3 === 1 ? 120 : 38
+    const file = `src/review/generated-${String(index).padStart(3, "0")}.ts`
+    const before = reviewSource(index, lines)
+    const after = before
+      .replace(`value_${index}_4`, `updated_${index}_4`)
+      .replace(
+        `value_${index}_${Math.max(8, Math.floor(lines / 2))}`,
+        `updated_${index}_${Math.max(8, Math.floor(lines / 2))}`,
+      )
+      .replace(`value_${index}_${lines - 4}`, `updated_${index}_${lines - 4}`)
+    return {
+      file,
+      patch: reviewPatch(file, before, after),
+      additions: 3,
+      deletions: 3,
+      status: "modified" as const,
+    }
+  })
+}
+
+function reviewSource(seed: number, lines: number) {
+  return Array.from(
+    { length: lines },
+    (_, index) => `export const value_${seed}_${index} = "${reviewWords(seed + index, index % 5 === 0 ? 180 : 42)}"`,
+  ).join("\n")
+}
+
+function reviewPatch(file: string, before: string, after: string) {
+  const beforeLines = before.split("\n")
+  const afterLines = after.split("\n")
+  return [
+    `diff --git a/${file} b/${file}`,
+    `--- a/${file}`,
+    `+++ b/${file}`,
+    `@@ -1,${beforeLines.length} +1,${afterLines.length} @@`,
+    ...beforeLines.flatMap((line, index) => {
+      const next = afterLines[index]!
+      if (line === next) return [` ${line}`]
+      return [`-${line}`, `+${next}`]
+    }),
+  ].join("\n")
+}
+
+function reviewWords(seed: number, length: number) {
+  const words = ["alpha", "bravo", "charlie", "delta", "echo", "foxtrot", "golf", "hotel", "india", "juliet"]
+  return Array.from({ length: Math.ceil(length / 7) }, (_, index) => words[(seed + index * 3) % words.length]).join(" ")
+}