chore(archive): park SWE-bench Benchmark (Beta) UI in .archive

.archive/README.md

@@ -135,3 +135,49 @@ Note that the repo's own `npm run check:unused-exports` does **not** find these.
 **Verification:** `tsc --noEmit` clean, full vitest suite green (701 files / 6390 tests), production webpack build clean.
 
 **To restore:** reverse the `git mv` for the file in question. No other edits are needed.
+
+---
+
+## SWE-bench "Benchmark (Beta)" UI — archived 2026-08-16
+
+The SWE-bench Pro benchmark runner UI (task browser, run builder, per-run
+session group in the chat panel, `benchmark` WorkStation tab) is parked. On
+`develop` it was already unreachable from normal UI: no menu offered the
+`benchmark` create target, nothing created a `benchmark` tab except the E2E
+seed helpers, and `BenchmarkTabSidebar` had no consumer. The only live entry
+was clicking a legacy "Benchmark run coordinator" session in the sidebar,
+which routed to the run-list surface. **Not** the Housekeeper _token_ benchmark
+(`housekeeperTokenBenchmark`, `integrations:housekeeper.benchmark.*`) — that
+is a different feature and stays live.
+
+**What moved here (self-contained to the feature):**
+
+- `src/features/BenchmarkPanel/` — panel, task selector, `useBenchmarkTasks`, `useBenchmarkAgentBatchRun`
+- `src/modules/WorkStation/shared/SidebarModules/Benchmark/` — `BenchmarkTabSidebar`
+- `src/modules/WorkStation/TabContent/renderers/benchmark.tsx` — the `benchmark` tab renderer
+- `src/engines/ChatPanel/panels/BenchmarkRunBuilder.tsx` — the run-builder creator surface
+- `src/store/benchmark/` — batch status / active batch atoms
+- `src/api/tauri/benchmark/` — the `benchmarkApi` client over the `benchmark_*` Tauri commands
+- `src/app/root/e2e/helpers/benchmark.ts` — E2E seed/inspect helpers
+- `tests/e2e/specs/core/{benchmark-run-ui,benchmark-docker-execution}.spec.mjs` (mirrored under `.archive/tests/`)
+
+**What deliberately stayed live:**
+
+- `src-tauri/src/benchmark/` — the Rust runner/commands still compile and are registered; they are now unreferenced from the frontend and can be removed in a backend-only PR
+- `src/config/agentIcons.tsx` `flask-conical` entry and `src/assets/fileTypeIcons/folder-benchmark*.svg` — generic icon registry / file-icon theme, not feature-specific
+- Housekeeper token benchmark (`src/modules/MainApp/Integrations/Housekeeper/HousekeeperCategoryView.tsx`, `rpc.validation.housekeeperTokenBenchmark`)
+
+**Shared files edited in place** to sever the branch:
+
+- `src/store/workstation/tabs/{types.ts,tabFactory.ts,storage.ts,index.ts,factories/{index,codeEditor}.ts}` — dropped `"benchmark"` from `WorkStationTabType`, its host mapping and persisted-type allow-list, and `BenchmarkTabData` / `benchmarkTabFactory` / `createBenchmarkTab`
+- `src/modules/WorkStation/TabContent/registry.ts`, `shared/SidebarModules/index.ts`, `shared/TabBar/components/SortableTab/index.tsx` (`BookLock` icon branch), `AppShell/CodeSidebarHeaderActions.tsx`
+- `src/store/ui/chatPanelAtom.ts`, `src/types/ui/chatPanel.ts`, `src/engines/ChatPanel/navigation/chatPanelSurfaceReducer.ts` — removed `CHAT_PANEL_CREATE_TARGET.BENCHMARK`, `CHAT_PANEL_CONTENT_MODE.BENCHMARK_SESSION_GROUP`, `CHAT_PANEL_SURFACE_KIND.BENCHMARK_SESSION_GROUP` and their navigate/reducer cases
+- `src/engines/ChatPanel/{index.tsx,ChatPanelContent.tsx,ChatPanelEmptyContent.tsx,hooks/useChatPanelContentState.tsx}` — removed the run-list mount, the run-builder creator branch, and `showBenchmarkSessionGroupContent`
+- `src/scaffold/NavigationSidebar/connectors/{useWorkstationSidebarHandlers.ts,useSessionMenuItems/{index.tsx,menuItemBuilders.tsx},WorkstationSidebarConnector/sidebarConnector.pinnedAndRevealData.ts}` — removed coordinator-session routing, child-session hiding, and master-row highlighting
+- `src/util/session/sessionDisplayMetadata.ts` — removed the `benchmark` flag / `flask-conical` icon override
+- `src/app/root/{E2EBootstrap.tsx,e2e/types.ts}`, `tests/e2e/wdio.conf.mjs` — removed helper wiring and the docker fixture builder
+- `src/i18n/locales/*/sessions.json` — removed `creator.benchmark.*` and `creator.createTarget.benchmark` (13 locales)
+
+**Behavior change:** legacy "Benchmark run coordinator" sessions and their child sessions now appear in the sidebar as ordinary sessions (previously the children were hidden and the coordinator opened the run list). Any persisted `benchmark` WorkStation tab is dropped by the storage allow-list on load.
+
+**To restore:** reverse the `git mv`s above and revert the in-place edits (see the archival commit).

src/app/root/E2EBootstrap.tsx

@@ -61,7 +61,6 @@ import {
   removeAccount,
 } from "./e2e/helpers/accounts";
 import { createAgentOrgHelpers } from "./e2e/helpers/agentOrgs";
-import { createBenchmarkE2EHelpers } from "./e2e/helpers/benchmark";
 import { createCloudHelpers } from "./e2e/helpers/cloud";
 import { createConfigHelpers } from "./e2e/helpers/config";
 import { createDebugEndpointHelpers } from "./e2e/helpers/debugEndpoints";
@@ -305,13 +304,6 @@ export const E2EBootstrap: FC = () => {
       openAgentStationDiff,
     } = createNavigationHelpers(store);
 
-    const {
-      seedBenchmarkRun,
-      inspectBenchmarkRun,
-      startLocalDockerBenchmarkRun,
-      getBenchmarkRunStatus,
-    } = createBenchmarkE2EHelpers(store);
-
     const {
       cloudSeedAuthState,
       cloudClearAuthState,
@@ -512,10 +504,6 @@ export const E2EBootstrap: FC = () => {
       openAgentTab,
       openOrgTab,
       inspectWorkstationSurface,
-      seedBenchmarkRun,
-      inspectBenchmarkRun,
-      startLocalDockerBenchmarkRun,
-      getBenchmarkRunStatus,
       cloudSeedAuthState,
       cloudClearAuthState,
       cloudReadAuthState,

src/app/root/e2e/types.ts

@@ -820,25 +820,6 @@ export interface E2EHelpers {
       agentConfigRootCount: number;
     }>
   >;
-  seedBenchmarkRun: (opts: {
-    batchId?: string;
-    sourcePath: string;
-    taskIds: string[];
-    activeTaskId?: string;
-  }) => Promise<Result<{ batchId: string; activeTaskId: string | null }>>;
-  inspectBenchmarkRun: () => Promise<
-    Result<{
-      batchStatus: Json | null;
-      activeBatchId: string | null;
-      activeTaskId: string | null;
-    }>
-  >;
-  startLocalDockerBenchmarkRun: (opts: {
-    sourcePath: string;
-    taskId: string;
-    patch: string;
-  }) => Promise<Result<{ status: Json }>>;
-  getBenchmarkRunStatus: (runId: string) => Promise<Result<{ status: Json }>>;
   seedUserPresence: (opts: {
     roles: CustomRoleDefinition[];
     presence: UserPresenceState;

src/engines/ChatPanel/ChatPanelContent.test.ts

@@ -9,10 +9,6 @@ vi.mock("./SessionContentView", () => ({
     createElement("div", { "data-gui-session": sessionId }),
 }));
 
-vi.mock("@src/features/BenchmarkPanel", () => ({
-  BenchmarkPanel: () => createElement("div", { "data-benchmark": "true" }),
-}));
-
 const { ChatPanelContent } = await import("./ChatPanelContent");
 
 function render(sessionViewMode: SessionViewMode): string {
@@ -25,7 +21,6 @@ function render(sessionViewMode: SessionViewMode): string {
       onSessionContinuation: () => undefined,
       paginationEnabled: false,
       position: "right" as const,
-      showBenchmarkSessionGroupContent: false,
       showPanelContent: true,
       showSessionContent: true,
       sessionViewMode,

src/engines/ChatPanel/ChatPanelContent.tsx

@@ -1,17 +1,11 @@
-import React, { Suspense } from "react";
+import React from "react";
 
 import type { SessionContinuation } from "@src/store/session/sessionTabPlacementAtom";
 import type { ChatHistoryDisplayMode } from "@src/store/ui/chatPanelAtom";
 
 import SessionContentView from "./SessionContentView";
 import type { SessionViewMode } from "./hooks/useSessionViewMode";
 
-const BenchmarkPanel = React.lazy(() =>
-  import("@src/features/BenchmarkPanel").then((module) => ({
-    default: module.BenchmarkPanel,
-  }))
-);
-
 interface ChatPanelContentProps {
   currentSessionId: string | null;
   emptyChatContent: React.ReactNode;
@@ -20,7 +14,6 @@ interface ChatPanelContentProps {
   displayMode: ChatHistoryDisplayMode;
   paginationEnabled: boolean;
   position: "left" | "right";
-  showBenchmarkSessionGroupContent: boolean;
   showPanelContent: boolean;
   showSessionContent: boolean;
   /** Non-GUI surface for the active session; mounted only while one is on. */
@@ -32,8 +25,8 @@ interface ChatPanelContentProps {
 }
 
 /**
- * The shared "chat column": session transcript, the benchmark run-list (still
- * contentMode-driven), and the Launchpad / creator surfaces (`emptyChatContent`).
+ * The shared "chat column": session transcript and the Launchpad / creator
+ * surfaces (`emptyChatContent`).
  * The workspace / organization / work-item / project / explore
  * surfaces are no longer rendered here — they are dedicated tab-typed renderers
  * dispatched by `UnifiedChatPanelTabContent`.
@@ -46,7 +39,6 @@ export function ChatPanelContent({
   displayMode,
   paginationEnabled,
   position,
-  showBenchmarkSessionGroupContent,
   showPanelContent,
   showSessionContent,
   alternateSessionView,
@@ -56,11 +48,7 @@ export function ChatPanelContent({
   const alternateActive = sessionViewMode !== "gui";
   return (
     <div className="relative flex min-h-0 min-w-0 flex-1 flex-col overflow-hidden">
-      {!showPanelContent ? null : showBenchmarkSessionGroupContent ? (
-        <Suspense fallback={null}>
-          <BenchmarkPanel surface="runList" />
-        </Suspense>
-      ) : showSessionContent && currentSessionId ? (
+      {!showPanelContent ? null : showSessionContent && currentSessionId ? (
         <>
           {/* Kept mounted while another view is showing: unmounting would drop
               the virtualized chat list's measurement cache and force a full

src/engines/ChatPanel/ChatPanelEmptyContent.tsx

@@ -41,11 +41,6 @@ const CreateWorkItemView = React.lazy(
   () =>
     import("@src/modules/ProjectManager/WorkItems/components/CreateWorkItemView")
 );
-const BenchmarkRunBuilder = React.lazy(() =>
-  import("./panels/BenchmarkRunBuilder").then((module) => ({
-    default: module.BenchmarkRunBuilder,
-  }))
-);
 
 type SessionCreatorSlot = NonNullable<ChatPanelProps["sessionCreatorSlot"]>;
 type SessionCreatorSlotProps = React.ComponentProps<SessionCreatorSlot>;
@@ -421,13 +416,5 @@ export function ChatPanelEmptyContent({
     return renderCollabOrgCreator();
   }
 
-  if (createTarget === CHAT_PANEL_CREATE_TARGET.BENCHMARK) {
-    return (
-      <Suspense fallback={null}>
-        <BenchmarkRunBuilder className={creatorClassName} />
-      </Suspense>
-    );
-  }
-
   return null;
 }

src/engines/ChatPanel/TabContent/UnifiedChatPanelTabContent.tsx

@@ -12,8 +12,8 @@ const WorkManagement = React.lazy(
 
 interface UnifiedChatPanelTabContentProps {
   activeTab: ChatPanelTab | null;
-  /** The shared "chat column" node (session transcript, Launchpad / creators,
-   *  and the benchmark run-list). Built by the host so this dispatcher stays
+  /** The shared "chat column" node (session transcript, Launchpad / creators).
+   *  Built by the host so this dispatcher stays
    *  agnostic of its heavy prop surface. */
   chatColumn: React.ReactNode;
   isTerminalTabActive: boolean;

src/engines/ChatPanel/TabContent/registry.ts

@@ -7,8 +7,8 @@
  * adding a new tab type is a type error until it has an entry — the same
  * guarantee the WorkStation `REGISTRY` provides.
  *
- * Session and Launchpad share the "chat column" (transcript / creators /
- * benchmark run-list, still contentMode-driven inside that column). Kanban and
+ * Session and Launchpad share the "chat column" (transcript / creators,
+ * still contentMode-driven inside that column). Kanban and
  * terminals render in their own keep-alive layers. Every other surface renders
  * a dedicated, self-sufficient component that reads its tab payload directly.
  */

src/engines/ChatPanel/hooks/useChatPanelContentState.tsx

@@ -21,7 +21,6 @@ interface UseChatPanelContentStateOptions {
 }
 
 export interface ChatPanelContentState {
-  showBenchmarkSessionGroupContent: boolean;
   showCloudOrgContent: boolean;
   showExploreContent: boolean;
   showExplicitNonSessionContent: boolean;
@@ -45,46 +44,33 @@ export function useChatPanelContentState({
   selectedWorkItem,
   selectedWorkspace,
 }: UseChatPanelContentStateOptions): ChatPanelContentState {
-  const showBenchmarkSessionGroupContent =
-    active && contentMode === CHAT_PANEL_CONTENT_MODE.BENCHMARK_SESSION_GROUP;
   const showSessionContent =
     active &&
-    !showBenchmarkSessionGroupContent &&
     contentMode === CHAT_PANEL_CONTENT_MODE.SESSION &&
     Boolean(currentSessionId);
-  const showWorkItemContent =
-    Boolean(selectedWorkItem) &&
-    !showBenchmarkSessionGroupContent &&
-    !showSessionContent;
+  const showWorkItemContent = Boolean(selectedWorkItem) && !showSessionContent;
   const showProjectContent =
-    Boolean(selectedProject) &&
-    !showBenchmarkSessionGroupContent &&
-    !showSessionContent &&
-    !showWorkItemContent;
+    Boolean(selectedProject) && !showSessionContent && !showWorkItemContent;
   const showProjectOrgContent =
     Boolean(selectedProjectOrg) &&
-    !showBenchmarkSessionGroupContent &&
     !showSessionContent &&
     !showWorkItemContent &&
     !showProjectContent;
   const showExploreContent =
     exploreOpen &&
-    !showBenchmarkSessionGroupContent &&
     !showSessionContent &&
     !showWorkItemContent &&
     !showProjectContent &&
     !showProjectOrgContent;
   const showCloudOrgContent =
     Boolean(selectedCloudOrg) &&
-    !showBenchmarkSessionGroupContent &&
     !showSessionContent &&
     !showWorkItemContent &&
     !showProjectContent &&
     !showProjectOrgContent &&
     !showExploreContent;
   const showWorkspaceOverviewContent =
     Boolean(selectedWorkspace) &&
-    !showBenchmarkSessionGroupContent &&
     !showSessionContent &&
     !showWorkItemContent &&
     !showProjectContent &&
@@ -95,7 +81,6 @@ export function useChatPanelContentState({
     contentMode === CHAT_PANEL_CONTENT_MODE.NON_SESSION;
   const showPanelContent =
     active ||
-    showBenchmarkSessionGroupContent ||
     showWorkItemContent ||
     showProjectContent ||
     showProjectOrgContent ||
@@ -104,7 +89,6 @@ export function useChatPanelContentState({
     showWorkspaceOverviewContent ||
     showExplicitNonSessionContent;
   const showHeader =
-    showBenchmarkSessionGroupContent ||
     showWorkItemContent ||
     showProjectContent ||
     showProjectOrgContent ||
@@ -115,7 +99,6 @@ export function useChatPanelContentState({
     active;
 
   return {
-    showBenchmarkSessionGroupContent,
     showCloudOrgContent,
     showExploreContent,
     showExplicitNonSessionContent,

src/engines/ChatPanel/index.tsx

@@ -660,9 +660,6 @@ const ChatPanel: React.FC<ChatPanelProps> = memo(
         onSessionContinuation={handleSessionContinuation}
         paginationEnabled={paginationEnabled}
         position={position}
-        showBenchmarkSessionGroupContent={
-          contentState.showBenchmarkSessionGroupContent
-        }
         showPanelContent={contentState.showPanelContent}
         showSessionContent={contentState.showSessionContent}
         sessionViewMode={sessionView.mode}

src/engines/ChatPanel/navigation/chatPanelSurfaceReducer.ts

@@ -57,11 +57,6 @@ export function reduceChatPanelSurfaceCommand(
         ...next,
         contentMode: CHAT_PANEL_CONTENT_MODE.SESSION,
       };
-    case CHAT_PANEL_SURFACE_KIND.BENCHMARK_SESSION_GROUP:
-      return {
-        ...next,
-        contentMode: CHAT_PANEL_CONTENT_MODE.BENCHMARK_SESSION_GROUP,
-      };
     case CHAT_PANEL_SURFACE_KIND.NEW_PROJECT:
       return {
         ...next,