You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@superset.apache.org by hu...@apache.org on 2026/07/21 16:17:06 UTC

(superset) branch master updated: feat(dashboard): export dashboard chart data to Excel (async, S3 + email) (#41133)

This is an automated email from the ASF dual-hosted git repository.

hughhhh pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/superset.git


The following commit(s) were added to refs/heads/master by this push:
     new 00cb3037e76 feat(dashboard): export dashboard chart data to Excel (async, S3 + email) (#41133)
00cb3037e76 is described below

commit 00cb3037e76ad304b07447ff6dc734e2b2ed4718
Author: Hugh A. Miles II <hu...@gmail.com>
AuthorDate: Tue Jul 21 12:17:00 2026 -0400

    feat(dashboard): export dashboard chart data to Excel (async, S3 + email) (#41133)
---
 UPDATING.md                                        |  25 ++
 docker/pythonpath_dev/superset_config.py           |   1 +
 .../using-superset/exporting-dashboard-data.mdx    |  99 +++++
 pyproject.toml                                     |   4 +
 .../DownloadMenuItems/DownloadMenuItems.test.tsx   | 148 +++++++-
 .../components/menu/DownloadMenuItems/index.tsx    |  71 +++-
 superset/charts/data/dashboard_filter_context.py   |  39 +-
 superset/config.py                                 |  22 ++
 superset/dashboards/api.py                         | 146 ++++++++
 superset/dashboards/excel_export/__init__.py       |  16 +
 superset/dashboards/excel_export/email.py          | 180 +++++++++
 superset/dashboards/excel_export/layout.py         |  97 +++++
 superset/dashboards/excel_export/screenshot.py     | 103 ++++++
 superset/dashboards/schemas.py                     |  29 +-
 superset/tasks/export_dashboard_excel.py           | 361 +++++++++++++++++++
 superset/utils/excel.py                            |   9 +-
 superset/utils/excel_streaming.py                  | 250 +++++++++++++
 superset/utils/s3.py                               |  83 +++++
 tests/integration_tests/dashboards/api_tests.py    | 169 +++++++++
 .../charts/test_dashboard_filter_context.py        | 172 +++++++++
 .../dashboards/test_excel_export_email.py          | 157 ++++++++
 .../dashboards/test_excel_export_layout.py         | 105 ++++++
 .../dashboards/test_excel_export_screenshot.py     | 164 +++++++++
 .../tasks/test_export_dashboard_excel.py           | 401 +++++++++++++++++++++
 tests/unit_tests/utils/excel_streaming_tests.py    | 267 ++++++++++++++
 tests/unit_tests/utils/s3_tests.py                 |  95 +++++
 26 files changed, 3193 insertions(+), 20 deletions(-)

diff --git a/UPDATING.md b/UPDATING.md
index e12c06540fc..4e829bcf8de 100644
--- a/UPDATING.md
+++ b/UPDATING.md
@@ -24,6 +24,31 @@ assists people when migrating to a new version.
 
 ## Next
 
+### Dashboard "Export Data to Excel" requires a Celery worker and S3 bucket
+
+A new dashboard action exports every chart's data to a single multi-sheet
+`.xlsx` asynchronously. It is disabled by default and turns on only when
+`EXCEL_EXPORT_S3_BUCKET` is set (the endpoint returns `501` otherwise). It also
+requires a running Celery worker and a configured SMTP transport, since the task
+emails the requesting user a pre-signed download link. New config keys:
+`EXCEL_EXPORT_S3_BUCKET`, `EXCEL_EXPORT_S3_KEY_PREFIX`,
+`EXCEL_EXPORT_LINK_TTL_SECONDS`, `EXCEL_EXPORT_S3_CLIENT_KWARGS`, and
+`EXCEL_EXPORT_TABLE_VIZ_TYPES`.
+
+The feature depends on `boto3`, which is **not** installed by default; install it
+with `pip install apache-superset[excel-export]`.
+
+A second mode, **Export Images to Excel**, embeds non-table charts as rendered
+images (which viz types stay tabular is controlled by
+`EXCEL_EXPORT_TABLE_VIZ_TYPES`). It renders through the headless webdriver, so the
+menu option only appears when the webdriver screenshot feature flags
+(`ENABLE_DASHBOARD_SCREENSHOT_ENDPOINTS`,
+`ENABLE_DASHBOARD_DOWNLOAD_WEBDRIVER_SCREENSHOT`) are enabled.
+
+Deployments that override `CELERY_CONFIG` must add
+`"superset.tasks.export_dashboard_excel"` to their `imports` tuple, or the task
+will not register and exports will silently never run.
+
 ### SQL_QUERY_MUTATOR now honors MUTATE_AFTER_SPLIT in SQL Lab
 
 SQL Lab now applies `SQL_QUERY_MUTATOR` according to `MUTATE_AFTER_SPLIT`, matching the documented semantics and the chart/query path. This only affects deployments that define `SQL_QUERY_MUTATOR` in `superset_config.py`:
diff --git a/docker/pythonpath_dev/superset_config.py b/docker/pythonpath_dev/superset_config.py
index ce7f1998708..7755cdb2cdc 100644
--- a/docker/pythonpath_dev/superset_config.py
+++ b/docker/pythonpath_dev/superset_config.py
@@ -87,6 +87,7 @@ class CeleryConfig:
         "superset.tasks.scheduler",
         "superset.tasks.thumbnails",
         "superset.tasks.cache",
+        "superset.tasks.export_dashboard_excel",
     )
     result_backend = f"redis://{REDIS_HOST}:{REDIS_PORT}/{REDIS_RESULTS_DB}"
     worker_prefetch_multiplier = 1
diff --git a/docs/docs/using-superset/exporting-dashboard-data.mdx b/docs/docs/using-superset/exporting-dashboard-data.mdx
new file mode 100644
index 00000000000..46c0a749d7c
--- /dev/null
+++ b/docs/docs/using-superset/exporting-dashboard-data.mdx
@@ -0,0 +1,99 @@
+---
+title: Exporting Dashboard Data to Excel
+hide_title: true
+sidebar_position: 7
+version: 1
+---
+
+# Exporting Dashboard Data to Excel
+
+Superset can export every chart on a dashboard to a single Excel workbook, with
+each chart's underlying data rendered as its own worksheet. The export reflects
+the dashboard's currently applied filters and runs asynchronously: when it
+finishes, the requesting user receives an email with a time-limited download
+link.
+
+## Using the export
+
+From a dashboard, open the **... (actions) → Download** submenu and choose
+**Export Data to Excel**. The action appears for users who have the dashboard
+`can_export` permission. You'll see a confirmation that the export is being
+prepared; the workbook arrives by email when it's ready.
+
+A second option, **Export Images to Excel**, embeds each non-table chart as a
+rendered image (tables stay tabular) instead of exporting raw data. Because it
+renders charts through the headless webdriver, this option only appears when the
+webdriver screenshot feature flags are enabled (see the prerequisites below);
+which viz types stay tabular is controlled by `EXCEL_EXPORT_TABLE_VIZ_TYPES`.
+
+Notes on the generated workbook:
+
+- One worksheet per chart, named `{chart_id} - {chart title}` (truncated to
+  Excel's 31-character limit; the chart id keeps names unique).
+- Charts nested in tabs are included.
+- Data reflects the dashboard's active filter state at the time of export.
+- A chart with no saved query context is skipped and listed in the email; open
+  the chart in Explore and re-save it to include it next time.
+- Row counts per sheet are capped the same way as the chart-level CSV/Excel
+  export (`ROW_LIMIT`, bounded by `SQL_MAX_ROW`), and never exceed Excel's
+  per-sheet maximum.
+
+## Prerequisites
+
+This feature is **disabled by default**. It requires:
+
+1. **The `boto3` dependency.** It is not installed by default; install it with
+   `pip install apache-superset[excel-export]`. Without it, exports fail and the
+   user receives a failure email.
+2. **An S3 bucket.** Set `EXCEL_EXPORT_S3_BUCKET`. Until it is set, the export
+   endpoint returns `501` and the menu action surfaces a "not configured"
+   message.
+3. **A running Celery worker.** The export runs as a Celery task. If no worker
+   is running, the request is accepted but nothing is produced.
+4. **A configured SMTP transport.** The download link is delivered by email
+   using the same settings as alerts & reports (`SMTP_*`,
+   `EMAIL_REPORTS_SUBJECT_PREFIX`).
+
+**Export Images to Excel** additionally requires a working headless webdriver —
+the same infrastructure scheduled reports and thumbnails use (`WEBDRIVER_*`,
+plus the `ENABLE_DASHBOARD_SCREENSHOT_ENDPOINTS` and
+`ENABLE_DASHBOARD_DOWNLOAD_WEBDRIVER_SCREENSHOT` feature flags). The menu option
+is hidden when those flags are off; if the webdriver is unreachable, image
+charts come back empty even though the data export path still works.
+
+Deployments that override `CELERY_CONFIG` must add
+`"superset.tasks.export_dashboard_excel"` to the `imports` tuple, or the task
+will not register.
+
+## Configuration keys
+
+| Key | Default | Description |
+| --- | --- | --- |
+| `EXCEL_EXPORT_S3_BUCKET` | `None` | Destination bucket. Required; `501` if unset. |
+| `EXCEL_EXPORT_S3_KEY_PREFIX` | `"dashboard-exports/"` | Key prefix: `{prefix}{dashboard_id}/{job_id}.xlsx`. |
+| `EXCEL_EXPORT_LINK_TTL_SECONDS` | `86400` | Lifetime of the pre-signed download URL (24h). |
+| `EXCEL_EXPORT_S3_CLIENT_KWARGS` | `{}` | Extra kwargs for `boto3.client("s3", ...)` — e.g. `region_name`, or `endpoint_url` for MinIO/LocalStack. |
+| `EXCEL_EXPORT_TABLE_VIZ_TYPES` | `None` | Viz types kept tabular in **Export Images to Excel** mode; every other type is embedded as an image. `None` uses the built-in default (`table`, `pivot_table`, `pivot_table_v2`). |
+
+Credentials and region resolve through the standard boto3 chain (environment
+variables, shared config, or instance role) unless overridden via
+`EXCEL_EXPORT_S3_CLIENT_KWARGS`. The worker needs `s3:PutObject` on the bucket.
+
+## Security considerations
+
+- The emailed link is a **pre-signed S3 URL**: anyone who holds it can download
+  the workbook until it expires. Keep the bucket **private**, enable
+  encryption, and consider a lifecycle rule to delete objects after a few days.
+  Lower `EXCEL_EXPORT_LINK_TTL_SECONDS` if 24 hours is too long for your data.
+- The export runs with the requesting user's permissions; each chart's query is
+  access-checked, so users only ever receive data they are entitled to.
+
+## Limitations
+
+- **Embedded dashboards / guest tokens are not supported** in this version,
+  because guest users have no email address to deliver the link to. Logged-in
+  users viewing an embedded dashboard can still use the export.
+- The default **Export Data to Excel** mode exports data only (no visual
+  styling). Use **Export Images to Excel** to embed rendered chart images, which
+  requires the webdriver infrastructure described in the prerequisites.
+- Scheduled/automated exports are not part of this feature.
diff --git a/pyproject.toml b/pyproject.toml
index 976b8d82ee3..1d3803c1fee 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -153,6 +153,10 @@ solr = ["sqlalchemy-solr >= 0.2.4.3"]
 elasticsearch = ["elasticsearch-dbapi>=0.2.13, <0.3.0"]
 exasol = ["sqlalchemy-exasol>=2.4.0, <8.0"]
 excel = ["xlrd>=2.0.2, <2.1"]
+# Async dashboard "Export Data/Images to Excel": uploads the workbook to S3 and
+# emails a pre-signed link. boto3 is imported lazily by superset.utils.s3, so
+# installing this extra is only required to actually run exports.
+excel-export = ["boto3"]
 fastmcp = [
     "fastmcp>=3.4.3,<4.0",
     # tiktoken backs the response-size-guard token estimator. Without
diff --git a/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/DownloadMenuItems.test.tsx b/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/DownloadMenuItems.test.tsx
index 86efc2860e0..c06cd6a06df 100644
--- a/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/DownloadMenuItems.test.tsx
+++ b/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/DownloadMenuItems.test.tsx
@@ -26,6 +26,7 @@ import {
 import { Menu, MenuItem } from '@superset-ui/core/components/Menu';
 import {
   FeatureFlag,
+  getClientErrorObject,
   isFeatureEnabled,
   SupersetClient,
 } from '@superset-ui/core';
@@ -46,12 +47,15 @@ jest.mock('src/components/MessageToasts/withToasts', () => ({
 jest.mock('@superset-ui/core', () => ({
   ...jest.requireActual('@superset-ui/core'),
   isFeatureEnabled: jest.fn().mockReturnValue(false),
+  getClientErrorObject: jest.fn().mockResolvedValue({}),
   SupersetClient: {
     get: jest.fn(),
+    post: jest.fn(),
   },
 }));
 
 const mockSupersetClient = SupersetClient as jest.Mocked<typeof SupersetClient>;
+const mockGetClientErrorObject = getClientErrorObject as jest.Mock;
 
 const createProps = () => ({
   pdfMenuItemTitle: 'Export to PDF',
@@ -70,19 +74,40 @@ const MenuWrapper = () => {
   return <Menu forceSubMenuRender items={menuItems} />;
 };
 
+const MenuWrapperWithProps = (
+  overrides: Partial<ReturnType<typeof createProps>> & {
+    canExportImage?: boolean;
+  },
+) => {
+  const downloadMenuItem = useDownloadMenuItems({
+    ...createProps(),
+    ...overrides,
+  });
+  const menuItems: MenuItem[] = [downloadMenuItem];
+  return <Menu forceSubMenuRender items={menuItems} />;
+};
+
 const originalCreateObjectURL = window.URL.createObjectURL;
 const originalRevokeObjectURL = window.URL.revokeObjectURL;
 
 beforeEach(() => {
   jest.clearAllMocks();
+  // Reset the implementation each test: clearAllMocks resets call history but
+  // not mockReturnValue, so an override in one test would otherwise leak.
+  (isFeatureEnabled as jest.Mock).mockReturnValue(false);
 });
 
+// "Export Images to Excel" is gated on the webdriver screenshot feature flags.
+const enableWebDriverScreenshot = () =>
+  (isFeatureEnabled as jest.Mock).mockReturnValue(true);
+
 afterEach(() => {
   window.URL.createObjectURL = originalCreateObjectURL;
   window.URL.revokeObjectURL = originalRevokeObjectURL;
 });
 
 test('Should render all menu items', () => {
+  enableWebDriverScreenshot();
   render(<MenuWrapper />, {
     useRedux: true,
   });
@@ -92,10 +117,120 @@ test('Should render all menu items', () => {
   expect(screen.getByText('Download as Image')).toBeInTheDocument();
 
   // Export options
+  expect(screen.getByText('Export Data to Excel')).toBeInTheDocument();
+  expect(screen.getByText('Export Images to Excel')).toBeInTheDocument();
   expect(screen.getByText('Export YAML')).toBeInTheDocument();
   expect(screen.getByText('Export as Example')).toBeInTheDocument();
 });
 
+test('Export Images to Excel is hidden when the webdriver is not enabled', () => {
+  // Default: webdriver screenshot flags off. Image export needs the webdriver,
+  // so only the data export is offered.
+  render(<MenuWrapper />, { useRedux: true });
+
+  expect(screen.getByText('Export Data to Excel')).toBeInTheDocument();
+  expect(screen.queryByText('Export Images to Excel')).not.toBeInTheDocument();
+});
+
+test('Excel export items are hidden when userCanExport is false', () => {
+  render(<MenuWrapperWithProps userCanExport={false} />, { useRedux: true });
+
+  expect(screen.queryByText('Export Data to Excel')).not.toBeInTheDocument();
+  expect(screen.queryByText('Export Images to Excel')).not.toBeInTheDocument();
+  // YAML export is not gated and remains visible
+  expect(screen.getByText('Export YAML')).toBeInTheDocument();
+});
+
+test('Export Data to Excel posts mode "data" and shows a pending toast', async () => {
+  mockSupersetClient.post.mockResolvedValue({
+    json: { job_id: 'abc' },
+  } as never);
+
+  render(<MenuWrapper />, { useRedux: true });
+
+  await userEvent.click(screen.getByText('Export Data to Excel'));
+
+  await waitFor(() => {
+    expect(mockSupersetClient.post).toHaveBeenCalledWith({
+      endpoint: '/api/v1/dashboard/123/export_xlsx/',
+      jsonPayload: { active_data_mask: {}, mode: 'data' },
+    });
+    expect(mockAddSuccessToast).toHaveBeenCalledWith(
+      "Your export is being prepared. You'll receive an email when it's ready.",
+    );
+  });
+});
+
+test('Export Images to Excel posts mode "images" and shows a pending toast', async () => {
+  enableWebDriverScreenshot();
+  mockSupersetClient.post.mockResolvedValue({
+    json: { job_id: 'abc' },
+  } as never);
+
+  render(<MenuWrapper />, { useRedux: true });
+
+  await userEvent.click(screen.getByText('Export Images to Excel'));
+
+  await waitFor(() => {
+    expect(mockSupersetClient.post).toHaveBeenCalledWith({
+      endpoint: '/api/v1/dashboard/123/export_xlsx/',
+      jsonPayload: { active_data_mask: {}, mode: 'images' },
+    });
+    expect(mockAddSuccessToast).toHaveBeenCalledWith(
+      "Your export is being prepared. You'll receive an email when it's ready.",
+    );
+  });
+});
+
+test('Export Data to Excel shows an "already in progress" toast when throttled', async () => {
+  // The throttle response is 202 with a message but no job_id.
+  mockSupersetClient.post.mockResolvedValue({
+    json: {
+      message: 'An Excel export for this dashboard is already in progress.',
+    },
+  } as never);
+
+  render(<MenuWrapper />, { useRedux: true });
+
+  await userEvent.click(screen.getByText('Export Data to Excel'));
+
+  await waitFor(() => {
+    expect(mockAddSuccessToast).toHaveBeenCalledWith(
+      'An export for this dashboard is already in progress.',
+    );
+  });
+});
+
+test('Export Data to Excel shows a config error toast on 501', async () => {
+  mockSupersetClient.post.mockRejectedValue(new Error('not configured'));
+  mockGetClientErrorObject.mockResolvedValue({ status: 501 });
+
+  render(<MenuWrapper />, { useRedux: true });
+
+  await userEvent.click(screen.getByText('Export Data to Excel'));
+
+  await waitFor(() => {
+    expect(mockAddDangerToast).toHaveBeenCalledWith(
+      'Excel export is not configured on this server.',
+    );
+  });
+});
+
+test('Export Data to Excel shows a generic error toast on other failures', async () => {
+  mockSupersetClient.post.mockRejectedValue(new Error('boom'));
+  mockGetClientErrorObject.mockResolvedValue({ status: 500 });
+
+  render(<MenuWrapper />, { useRedux: true });
+
+  await userEvent.click(screen.getByText('Export Data to Excel'));
+
+  await waitFor(() => {
+    expect(mockAddDangerToast).toHaveBeenCalledWith(
+      'Sorry, something went wrong. Try again later.',
+    );
+  });
+});
+
 test('Export as Example calls SupersetClient.get with correct endpoint', async () => {
   const mockBlob = new Blob(['test'], { type: 'application/zip' });
   const mockResponse: Pick<Response, 'blob' | 'headers'> = {
@@ -144,19 +279,6 @@ test('Export as Example shows error toast on failure', async () => {
 
 const mockIsFeatureEnabled = isFeatureEnabled as jest.Mock;
 
-const MenuWrapperWithProps = (
-  overrides: Partial<ReturnType<typeof createProps>> & {
-    canExportImage?: boolean;
-  },
-) => {
-  const downloadMenuItem = useDownloadMenuItems({
-    ...createProps(),
-    ...overrides,
-  });
-  const menuItems: MenuItem[] = [downloadMenuItem];
-  return <Menu forceSubMenuRender items={menuItems} />;
-};
-
 test('Screenshot menu items should be disabled when GranularExportControls is ON and canExportImage is false', () => {
   mockIsFeatureEnabled.mockImplementation(
     (flag: string) => flag === FeatureFlag.GranularExportControls,
diff --git a/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/index.tsx b/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/index.tsx
index 0379588bf7a..248d2795260 100644
--- a/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/index.tsx
+++ b/superset-frontend/src/dashboard/components/menu/DownloadMenuItems/index.tsx
@@ -17,17 +17,20 @@
  * under the License.
  */
 import { SyntheticEvent } from 'react';
+import { useSelector } from 'react-redux';
 import { logging } from '@apache-superset/core/utils';
 import { t } from '@apache-superset/core/translation';
 import {
   FeatureFlag,
+  getClientErrorObject,
   isFeatureEnabled,
   SupersetClient,
 } from '@superset-ui/core';
 import { MenuItem } from '@superset-ui/core/components/Menu';
 import { parse as parseContentDisposition } from 'content-disposition';
 import { useDownloadScreenshot } from 'src/dashboard/hooks/useDownloadScreenshot';
-import { MenuKeys } from 'src/dashboard/types';
+import { NATIVE_FILTER_PREFIX } from 'src/dashboard/components/nativeFilters/FiltersConfigModal/utils';
+import { MenuKeys, RootState } from 'src/dashboard/types';
 import downloadAsPdf from 'src/utils/downloadAsPdf';
 import downloadAsImage from 'src/utils/downloadAsImage';
 import handleResourceExport from 'src/utils/export';
@@ -68,8 +71,19 @@ export const useDownloadMenuItems = (
   } = props;
 
   const { addDangerToast, addSuccessToast } = useToasts();
+  const dataMask = useSelector((state: RootState) => state.dataMask);
   const SCREENSHOT_NODE_SELECTOR = '.dashboard';
 
+  const buildActiveDataMask = (): Record<string, { extraFormData: object }> =>
+    Object.entries(dataMask || {}).reduce<
+      Record<string, { extraFormData: object }>
+    >((acc, [id, mask]) => {
+      if (id.startsWith(NATIVE_FILTER_PREFIX)) {
+        acc[id] = { extraFormData: mask?.extraFormData ?? {} };
+      }
+      return acc;
+    }, {});
+
   const isWebDriverScreenshotEnabled =
     isFeatureEnabled(FeatureFlag.EnableDashboardScreenshotEndpoints) &&
     isFeatureEnabled(FeatureFlag.EnableDashboardDownloadWebDriverScreenshot);
@@ -153,6 +167,39 @@ export const useDownloadMenuItems = (
     }
   };
 
+  const onExportXlsx = async (mode: 'data' | 'images') => {
+    try {
+      const { json } = await SupersetClient.post({
+        endpoint: `/api/v1/dashboard/${dashboardId}/export_xlsx/`,
+        jsonPayload: { active_data_mask: buildActiveDataMask(), mode },
+      });
+      // The throttle response (an export is already running) returns 202 with a
+      // message but no job_id; only a freshly enqueued job carries a job_id.
+      if ((json as { job_id?: string })?.job_id) {
+        addSuccessToast(
+          t(
+            "Your export is being prepared. You'll receive an email when it's ready.",
+          ),
+        );
+      } else {
+        addSuccessToast(
+          t('An export for this dashboard is already in progress.'),
+        );
+      }
+    } catch (error) {
+      // status comes from the response (Partial<SupersetClientResponse>), which
+      // the union type does not expose uniformly; read it via a narrow cast.
+      const { status } = (await getClientErrorObject(error)) as {
+        status?: number;
+      };
+      if (status === 501) {
+        addDangerToast(t('Excel export is not configured on this server.'));
+      } else {
+        addDangerToast(t('Sorry, something went wrong. Try again later.'));
+      }
+    }
+  };
+
   const imageDisabled = canExportImage === false;
 
   const imageExportLabel = (text: string) =>
@@ -198,6 +245,28 @@ export const useDownloadMenuItems = (
       ];
 
   const exportMenuItems: MenuItem[] = [
+    ...(userCanExport
+      ? [
+          {
+            key: 'export-xlsx',
+            label: t('Export Data to Excel'),
+            onClick: () => onExportXlsx('data'),
+          },
+          // Image export renders charts through the headless webdriver, so only
+          // offer it where that infrastructure is available (same signal as the
+          // PDF/PNG image downloads above); otherwise non-table charts would
+          // silently come back empty.
+          ...(isWebDriverScreenshotEnabled
+            ? [
+                {
+                  key: 'export-xlsx-images',
+                  label: t('Export Images to Excel'),
+                  onClick: () => onExportXlsx('images'),
+                },
+              ]
+            : []),
+        ]
+      : []),
     {
       key: 'export-yaml',
       label: t('Export YAML'),
diff --git a/superset/charts/data/dashboard_filter_context.py b/superset/charts/data/dashboard_filter_context.py
index bd36bd07432..1eda538b6fc 100644
--- a/superset/charts/data/dashboard_filter_context.py
+++ b/superset/charts/data/dashboard_filter_context.py
@@ -199,6 +199,31 @@ def _extract_filter_extra_form_data(
     return None, DashboardFilterStatus.NOT_APPLIED
 
 
+def _resolve_filter_extra_form_data(
+    filter_config: dict[str, Any],
+    active_data_mask: dict[str, Any] | None,
+) -> tuple[dict[str, Any] | None, DashboardFilterStatus]:
+    """
+    Resolve a filter's extra_form_data and status, preferring an active value
+    from ``active_data_mask`` over the filter's saved default.
+
+    When ``active_data_mask`` provides an entry for this filter, its
+    ``extraFormData`` is authoritative: a non-empty value is APPLIED, while an
+    empty value means the user explicitly cleared the filter (NOT_APPLIED, with
+    no fallback to the saved default). When no active entry exists, fall back to
+    the saved-default behavior in ``_extract_filter_extra_form_data``.
+
+    Returns (extra_form_data, status).
+    """
+    flt_id = filter_config.get("id", "")
+    if active_data_mask is not None and flt_id in active_data_mask:
+        active_efd = (active_data_mask[flt_id] or {}).get("extraFormData") or {}
+        if active_efd:
+            return active_efd, DashboardFilterStatus.APPLIED
+        return None, DashboardFilterStatus.NOT_APPLIED
+    return _extract_filter_extra_form_data(filter_config)
+
+
 def _get_filter_target_column(filter_config: dict[str, Any]) -> str | None:
     """Extract the target column name from a native filter configuration."""
     if targets := filter_config.get("targets", []):
@@ -244,16 +269,26 @@ def _check_dashboard_access(dashboard: Dashboard) -> None:
 def get_dashboard_filter_context(
     dashboard_id: int,
     chart_id: int,
+    *,
+    active_data_mask: dict[str, Any] | None = None,
 ) -> DashboardFilterContext:
     """
     Build a DashboardFilterContext for a chart on a dashboard.
 
     Loads the dashboard's native filter configuration, determines which
-    filters are in scope for the given chart, extracts default filter values,
+    filters are in scope for the given chart, resolves each filter's value,
     and returns the merged extra_form_data along with metadata about each filter.
 
+    When ``active_data_mask`` is provided (e.g. the live filter state from a
+    dashboard view), each in-scope filter present in the mask uses its active
+    ``extraFormData`` instead of the saved default; an empty active value means
+    the filter was cleared. Filters absent from the mask fall back to their
+    saved defaults, so omitting ``active_data_mask`` reproduces the dashboard's
+    initial-load behavior.
+
     :param dashboard_id: The ID of the dashboard
     :param chart_id: The ID of the chart
+    :param active_data_mask: Optional live filter state keyed by native filter id
     :returns: DashboardFilterContext with merged extra_form_data and filter metadata
     :raises ValueError: if dashboard not found or chart not on dashboard
     :raises SupersetSecurityException: if the user cannot access the dashboard
@@ -287,7 +322,7 @@ def get_dashboard_filter_context(
         flt_id = flt.get("id", "")
         flt_name = flt.get("name", "")
         target_column = _get_filter_target_column(flt)
-        extra_form_data, status = _extract_filter_extra_form_data(flt)
+        extra_form_data, status = _resolve_filter_extra_form_data(flt, active_data_mask)
 
         if extra_form_data and status == DashboardFilterStatus.APPLIED:
             context.extra_form_data = _merge_extra_form_data(
diff --git a/superset/config.py b/superset/config.py
index a745f4f234a..bfee0039947 100644
--- a/superset/config.py
+++ b/superset/config.py
@@ -1443,6 +1443,27 @@ CSV_STREAMING_ROW_THRESHOLD = 100000
 # note: index option should not be overridden
 EXCEL_EXPORT: dict[str, Any] = {}
 
+# ---------------------------------------------------
+# Dashboard "Export Data to Excel" (async, S3-backed)
+# ---------------------------------------------------
+# Destination S3 bucket for generated dashboard .xlsx exports. The feature is
+# disabled until this is set: the export endpoint returns 501 when it is None.
+EXCEL_EXPORT_S3_BUCKET: str | None = None
+# Key prefix for export objects: {prefix}{dashboard_id}/{job_id}.xlsx
+EXCEL_EXPORT_S3_KEY_PREFIX = "dashboard-exports/"
+# Lifetime (seconds) of the pre-signed download URL emailed to the user (24h).
+# Note: AWS S3 caps pre-signed URL lifetime at 7 days (604800 seconds); larger
+# values are rejected by S3, so keep this at or below that when using AWS.
+EXCEL_EXPORT_LINK_TTL_SECONDS = 86400
+# Extra kwargs passed to boto3.client("s3", ...) — e.g. region_name, or an
+# endpoint_url for S3-compatible stores (MinIO/LocalStack). Credentials
+# otherwise resolve through the standard boto3 chain.
+EXCEL_EXPORT_S3_CLIENT_KWARGS: dict[str, Any] = {}
+# Viz types treated as tables in the "Export Images to Excel" mode: these charts
+# stay tabular (one worksheet of data) while every other viz type is embedded as
+# a rendered image. Set to None to fall back to the built-in default.
+EXCEL_EXPORT_TABLE_VIZ_TYPES: set[str] | None = None
+
 # ---------------------------------------------------
 # Time grain configurations
 # ---------------------------------------------------
@@ -1635,6 +1656,7 @@ class CeleryConfig:  # pylint: disable=too-few-public-methods
         "superset.tasks.thumbnails",
         "superset.tasks.cache",
         "superset.tasks.slack",
+        "superset.tasks.export_dashboard_excel",
     )
     result_backend = "db+sqlite:///celery_results.sqlite"
     worker_prefetch_multiplier = 1
diff --git a/superset/dashboards/api.py b/superset/dashboards/api.py
index 45fb970a57b..5a138d0d401 100644
--- a/superset/dashboards/api.py
+++ b/superset/dashboards/api.py
@@ -17,6 +17,7 @@
 # pylint: disable=too-many-lines
 import functools
 import logging
+import uuid
 from datetime import datetime
 from io import BytesIO
 from typing import Any, Callable, cast
@@ -82,6 +83,8 @@ from superset.commands.dashboard.update import (
     UpdateDashboardNativeFiltersCommand,
 )
 from superset.commands.database.exceptions import DatasetValidationError
+from superset.commands.distributed_lock.acquire import AcquireDistributedLock
+from superset.commands.distributed_lock.release import ReleaseDistributedLock
 from superset.commands.exceptions import TagForbiddenError
 from superset.commands.importers.exceptions import NoValidFilesFoundError
 from superset.commands.importers.v1.utils import get_contents_from_bundle
@@ -107,6 +110,8 @@ from superset.dashboards.schemas import (
     DashboardColorsConfigUpdateSchema,
     DashboardCopySchema,
     DashboardDatasetSchema,
+    DashboardExportXlsxPostSchema,
+    DashboardExportXlsxResponseSchema,
     DashboardGetResponseSchema,
     DashboardNativeFiltersConfigUpdateSchema,
     DashboardPostSchema,
@@ -124,7 +129,9 @@ from superset.dashboards.schemas import (
     thumbnail_query_schema,
 )
 from superset.exceptions import (
+    LockAlreadyHeldException,
     ScreenshotImageNotAvailableException,
+    SupersetSecurityException,
 )
 from superset.extensions import event_logger, security_manager
 from superset.models.dashboard import Dashboard
@@ -135,6 +142,12 @@ from superset.subjects.filters import (
     FilterRelatedSubjects,
     subject_type_filter,
 )
+from superset.tasks.export_dashboard_excel import (
+    export_dashboard_excel,
+    EXPORT_LOCK_NAMESPACE,
+    export_lock_params,
+    EXPORT_LOCK_TTL_SECONDS,
+)
 from superset.tasks.thumbnails import (
     cache_dashboard_screenshot,
     cache_dashboard_thumbnail,
@@ -274,6 +287,7 @@ class DashboardRestApi(
         "put_chart_customizations",
         "put_colors",
         "export_as_example",
+        "export_xlsx",
         "list_versions",
         "get_version",
         "activity",
@@ -292,6 +306,10 @@ class DashboardRestApi(
     method_permission_name = {
         **MODEL_API_RW_METHOD_PERMISSION_MAP,
         "restore": "write",
+        # Reuse the dashboard ``can_export`` permission (the frontend gates the
+        # menu item on it) instead of the ``can_export_xlsx`` FAB would otherwise
+        # derive from the method name.
+        "export_xlsx": "export",
     }
 
     # Default list_columns (used if config not set)
@@ -497,6 +515,8 @@ class DashboardRestApi(
         DashboardCopySchema,
         DashboardGetResponseSchema,
         DashboardDatasetSchema,
+        DashboardExportXlsxPostSchema,
+        DashboardExportXlsxResponseSchema,
         TabsPayloadSchema,
         GetFavStarIdsSchema,
         EmbeddedDashboardResponseSchema,
@@ -1575,6 +1595,132 @@ class DashboardRestApi(
             response.set_cookie(token, "done", max_age=600)
         return response
 
+    @expose("/<pk>/export_xlsx/", methods=("POST",))
+    @protect()
+    @safe
+    @statsd_metrics
+    @event_logger.log_this_with_context(
+        action=lambda self, *args, **kwargs: f"{self.__class__.__name__}.export_xlsx",
+        log_to_statsd=False,
+    )
+    def export_xlsx(self, pk: int) -> WerkzeugResponse:
+        """Export all of a dashboard's chart data to an Excel workbook (async).
+        ---
+        post:
+          summary: Export dashboard chart data to Excel
+          description: >-
+            Enqueues an async task that writes each chart's data to its own
+            worksheet, uploads the .xlsx to S3, and emails the requesting user a
+            pre-signed download link. Returns immediately with a job id.
+          parameters:
+          - in: path
+            schema:
+              type: integer
+            name: pk
+            description: The dashboard id
+          requestBody:
+            content:
+              application/json:
+                schema:
+                  $ref: '#/components/schemas/DashboardExportXlsxPostSchema'
+          responses:
+            202:
+              description: Export task accepted
+              content:
+                application/json:
+                  schema:
+                    $ref: '#/components/schemas/DashboardExportXlsxResponseSchema'
+            400:
+              $ref: '#/components/responses/400'
+            401:
+              $ref: '#/components/responses/401'
+            403:
+              $ref: '#/components/responses/403'
+            404:
+              $ref: '#/components/responses/404'
+            500:
+              $ref: '#/components/responses/500'
+            501:
+              description: Excel export is not configured on this server
+        """
+        if not current_app.config["EXCEL_EXPORT_S3_BUCKET"]:
+            return self.response(
+                501, message="Excel export is not configured on this server."
+            )
+        try:
+            # Tolerate an empty/non-JSON body (e.g. a POST with no Content-Type);
+            # request.json would otherwise raise 415.
+            payload = DashboardExportXlsxPostSchema().load(
+                request.get_json(silent=True) or {}
+            )
+        except ValidationError as error:
+            return self.response_400(message=error.messages)
+
+        # Image export drives the headless webdriver, so it is only available
+        # when the same screenshot flags the UI checks are enabled. The decorator
+        # form (``@validate_feature_flags``) can't be used here because it would
+        # also block ``mode="data"``; mirror its 404 behavior inline instead.
+        if payload.get("mode") == "images" and not (
+            is_feature_enabled("ENABLE_DASHBOARD_SCREENSHOT_ENDPOINTS")
+            and is_feature_enabled("ENABLE_DASHBOARD_DOWNLOAD_WEBDRIVER_SCREENSHOT")
+        ):
+            return self.response_404()
+
+        dashboard = cast(Dashboard, self.datamodel.get(pk, self._base_filters))
+        if not dashboard:
+            return self.response_404()
+        try:
+            security_manager.raise_for_access(dashboard=dashboard)
+        except SupersetSecurityException:
+            return self.response_403()
+
+        # Email delivery is the only result channel, so an account with an email
+        # address is required; embedded guest users are excluded in this version.
+        if isinstance(g.user, GuestUser) or not getattr(g.user, "email", None):
+            return self.response_400(
+                message="Excel export requires an account with an email address."
+            )
+        if not dashboard.slices:
+            return self.response_400(message="Dashboard has no charts to export.")
+
+        # Throttle: one concurrent export per user+dashboard. Acquire a shared,
+        # atomic distributed lock (Redis when configured, the metadata DB
+        # otherwise) so the guard works across the web server and workers and is
+        # not a no-op under the default cache. The task releases it when it
+        # settles; the TTL is the backstop if that release is ever lost.
+        lock_params = export_lock_params(g.user.id, dashboard.id)
+        try:
+            AcquireDistributedLock(
+                EXPORT_LOCK_NAMESPACE,
+                lock_params,
+                ttl_seconds=EXPORT_LOCK_TTL_SECONDS,
+            ).run()
+        except LockAlreadyHeldException:
+            return self.response(
+                202,
+                message="An Excel export for this dashboard is already in progress.",
+            )
+
+        job_id = str(uuid.uuid4())
+        try:
+            export_dashboard_excel.apply_async(
+                kwargs={
+                    "dashboard_id": dashboard.id,
+                    "user_id": g.user.id,
+                    "active_data_mask": payload.get("active_data_mask", {}),
+                    "job_id": job_id,
+                    "mode": payload.get("mode", "data"),
+                },
+                task_id=job_id,
+            )
+        except Exception:
+            # If enqueuing fails (e.g. broker down) the task will never run to
+            # release the lock, so free it now rather than block exports until
+            # the TTL expires.
+            ReleaseDistributedLock(EXPORT_LOCK_NAMESPACE, lock_params).run()
+            raise
+        return self.response(202, job_id=job_id)
+
     @expose("/<pk>/cache_dashboard_screenshot/", methods=("POST",))
     @validate_feature_flags(["THUMBNAILS", "ENABLE_DASHBOARD_SCREENSHOT_ENDPOINTS"])
     @protect()
diff --git a/superset/dashboards/excel_export/__init__.py b/superset/dashboards/excel_export/__init__.py
new file mode 100644
index 00000000000..13a83393a91
--- /dev/null
+++ b/superset/dashboards/excel_export/__init__.py
@@ -0,0 +1,16 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
diff --git a/superset/dashboards/excel_export/email.py b/superset/dashboards/excel_export/email.py
new file mode 100644
index 00000000000..19233c0ac34
--- /dev/null
+++ b/superset/dashboards/excel_export/email.py
@@ -0,0 +1,180 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+Email rendering and delivery for dashboard Excel exports.
+
+Bodies use inline styles only (no external CSS, no logo) to match Superset's
+existing report notification emails, and all user-controlled values (dashboard
+title, chart names) are HTML-escaped to avoid injection.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime
+
+from flask import current_app
+from flask_babel import gettext as __, ngettext
+from markupsafe import escape
+
+from superset.utils.core import send_email_smtp
+
+_DATETIME_FORMAT = "%Y-%m-%d %H:%M:%S"
+_FOOTER_STYLE = "color:#888;font-size:12px;"
+_BUTTON_STYLE = (
+    "display:inline-block;padding:10px 16px;background:#20a7c9;color:#ffffff;"
+    "text-decoration:none;border-radius:4px;"
+)
+
+# Reason keys under which the export task groups charts it could not export.
+# The task classifies each omitted chart under one of these; the email renders a
+# separate, labelled section per non-empty group with its own remediation text.
+ERROR_NO_QUERY_CONTEXT = "no-query-context"
+ERROR_GENERAL = "general-exception"
+
+
+def _fmt(dt: datetime) -> str:
+    return dt.strftime(_DATETIME_FORMAT)
+
+
+def _humanize_ttl(seconds: int) -> str:
+    """Render a TTL as a human-readable, pluralized, translatable duration.
+
+    Whole hours read as "24 hours"; sub-hour and non-hour values keep their
+    minutes (e.g. "1 hour 30 minutes", "15 minutes") so the stated lifetime
+    always matches the real pre-signed URL expiration.
+    """
+    hours, remainder = divmod(seconds, 3600)
+    parts: list[str] = []
+    if hours:
+        parts.append(ngettext("%(num)d hour", "%(num)d hours", hours))
+    if minutes := remainder // 60:
+        parts.append(ngettext("%(num)d minute", "%(num)d minutes", minutes))
+    if not parts:
+        parts.append(ngettext("%(num)d second", "%(num)d seconds", seconds))
+    return " ".join(parts)
+
+
+def build_subject(dashboard_title: str, *, success: bool) -> str:
+    """Build the email subject, prefixed with EMAIL_REPORTS_SUBJECT_PREFIX."""
+    prefix = current_app.config["EMAIL_REPORTS_SUBJECT_PREFIX"]
+    if success:
+        return prefix + __(
+            "Your dashboard export is ready: %(title)s", title=dashboard_title
+        )
+    return prefix + __(
+        "Your dashboard export could not be completed: %(title)s",
+        title=dashboard_title,
+    )
+
+
+def _errored_section(errored: dict[str, list[str]]) -> str:
+    """Render one labelled, translated sub-list per non-empty error group.
+
+    ``errored`` maps a reason key (see the ``ERROR_*`` constants) to the labels
+    of the charts that were omitted for that reason. Known reasons are rendered
+    first, in a stable order, each with its own remediation text; any unknown
+    reason key falls back to a generic message so nothing is silently dropped.
+    """
+    if not errored:
+        return ""
+    notes = {
+        ERROR_NO_QUERY_CONTEXT: __(
+            "The following charts were omitted because they have no saved query "
+            "context. To include them, open each chart in Explore and re-save."
+        ),
+        ERROR_GENERAL: __(
+            "The following charts were omitted because an error occurred while "
+            "exporting them:"
+        ),
+    }
+    fallback = __("The following charts could not be exported:")
+    ordered = [ERROR_NO_QUERY_CONTEXT, ERROR_GENERAL]
+    reasons = ordered + [reason for reason in errored if reason not in ordered]
+    sections = []
+    for reason in reasons:
+        labels = errored.get(reason)
+        if not labels:
+            continue
+        note = notes.get(reason, fallback)
+        items = "".join(f"<li>{escape(label)}</li>" for label in labels)
+        sections.append(f"<p>{note}</p><ul>{items}</ul>")
+    return "".join(sections)
+
+
+def build_success_email(
+    dashboard_title: str,
+    download_url: str,
+    requested_at: datetime,
+    expires_at: datetime,
+    ttl_seconds: int,
+    errored: dict[str, list[str]],
+) -> str:
+    """Render the success email body (HTML)."""
+    title = escape(dashboard_title)
+    url = escape(download_url)
+    ready = __('Your export of "%(title)s" is ready.', title=title)
+    button = __("Download Excel file")
+    expiry = __(
+        "This link expires in %(duration)s (%(when)s UTC).",
+        duration=_humanize_ttl(ttl_seconds),
+        when=_fmt(expires_at),
+    )
+    requested = __(
+        "This export was requested on %(when)s UTC.", when=_fmt(requested_at)
+    )
+    disclaimer = __("If you did not request this, you can ignore this email.")
+    return (
+        '<html><body style="font-family:Arial,sans-serif;color:#333;">'
+        f"<p>{ready}</p>"
+        f'<p><a href="{url}" style="{_BUTTON_STYLE}">{button}</a></p>'
+        f"<p>{expiry}</p>"
+        f"{_errored_section(errored)}"
+        "<hr/>"
+        f'<p style="{_FOOTER_STYLE}">{requested}<br/>{disclaimer}</p>'
+        "</body></html>"
+    )
+
+
+def build_failure_email(dashboard_title: str, requested_at: datetime) -> str:
+    """Render the failure email body (HTML)."""
+    title = escape(dashboard_title)
+    failed = __('Your export of "%(title)s" could not be completed.', title=title)
+    advice = __(
+        "An error occurred while generating the file. Please try again, or "
+        "contact your administrator if the problem persists."
+    )
+    requested = __(
+        "This export was requested on %(when)s UTC.", when=_fmt(requested_at)
+    )
+    return (
+        '<html><body style="font-family:Arial,sans-serif;color:#333;">'
+        f"<p>{failed}</p>"
+        f"<p>{advice}</p>"
+        "<hr/>"
+        f'<p style="{_FOOTER_STYLE}">{requested}</p>'
+        "</body></html>"
+    )
+
+
+def send_export_email(to: str, subject: str, html_content: str) -> None:
+    """Send an export email via the configured SMTP transport."""
+    send_email_smtp(
+        to=to,
+        subject=subject,
+        html_content=html_content,
+        config=current_app.config,
+    )
diff --git a/superset/dashboards/excel_export/layout.py b/superset/dashboards/excel_export/layout.py
new file mode 100644
index 00000000000..a18299b2fb3
--- /dev/null
+++ b/superset/dashboards/excel_export/layout.py
@@ -0,0 +1,97 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""Determine the order in which a dashboard's charts appear in its layout."""
+
+from __future__ import annotations
+
+from typing import Any, TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from superset.models.dashboard import Dashboard
+    from superset.models.slice import Slice
+
+CHART_TYPE = "CHART"
+ROOT_ID = "ROOT_ID"
+
+
+def _walk_chart_ids(position: dict[str, Any]) -> list[int]:
+    """
+    Depth-first walk of a dashboard ``position_json`` returning chart ids in
+    visual (layout) order, including tab-nested charts. Each chart id appears
+    once (first occurrence wins); cycles are guarded against.
+    """
+    if ROOT_ID not in position:
+        return []
+
+    ordered: list[int] = []
+    seen_charts: set[int] = set()
+    visited_nodes: set[str] = set()
+    stack: list[str] = [ROOT_ID]
+
+    while stack:
+        node_id = stack.pop()
+        if node_id in visited_nodes:
+            continue
+        visited_nodes.add(node_id)
+
+        node = position.get(node_id)
+        if not isinstance(node, dict):
+            continue
+
+        if node.get("type") == CHART_TYPE:
+            chart_id = node.get("meta", {}).get("chartId")
+            if isinstance(chart_id, int) and chart_id not in seen_charts:
+                seen_charts.add(chart_id)
+                ordered.append(chart_id)
+
+        # Push children in reverse so they are popped in their declared order.
+        children = node.get("children", [])
+        for child_id in reversed(children):
+            stack.append(child_id)
+
+    return ordered
+
+
+def get_charts_in_layout_order(dashboard: Dashboard) -> list[Slice]:
+    """
+    Return the dashboard's charts ordered by their position in the layout.
+
+    Charts are visited depth-first over ``position_json`` (so tab-nested charts
+    are included in tab order), de-duplicated when the same chart is placed more
+    than once, and any chart that belongs to the dashboard but is absent from
+    the layout is appended at the end ordered by id. Layout entries that no
+    longer correspond to a dashboard chart are skipped.
+
+    :param dashboard: The dashboard whose charts to order
+    :returns: The dashboard's :class:`Slice` objects in layout order
+    """
+    slices_by_id: dict[int, Slice] = {slc.id: slc for slc in dashboard.slices}
+
+    result: list[Slice] = []
+    used: set[int] = set()
+    for chart_id in _walk_chart_ids(dashboard.position):
+        slc = slices_by_id.get(chart_id)
+        if slc is not None and chart_id not in used:
+            used.add(chart_id)
+            result.append(slc)
+
+    orphans = sorted(
+        (slc for chart_id, slc in slices_by_id.items() if chart_id not in used),
+        key=lambda slc: slc.id,
+    )
+    result.extend(orphans)
+    return result
diff --git a/superset/dashboards/excel_export/screenshot.py b/superset/dashboards/excel_export/screenshot.py
new file mode 100644
index 00000000000..c6c30759fa8
--- /dev/null
+++ b/superset/dashboards/excel_export/screenshot.py
@@ -0,0 +1,103 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+Render a single dashboard chart to a PNG for the image-mode Excel export.
+
+This reuses the same headless render path scheduled reports use
+(:class:`~superset.utils.screenshots.ChartScreenshot`), but points it at an
+Explore URL whose ``form_data`` carries the live dashboard filter state — so an
+embedded image reflects the same filters the data path applies, rather than the
+chart's default saved state.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+from celery.exceptions import SoftTimeLimitExceeded
+from flask import current_app
+
+from superset.charts.data.dashboard_filter_context import (
+    get_dashboard_filter_context,
+)
+from superset.utils import json
+from superset.utils.screenshots import ChartScreenshot
+from superset.utils.urls import get_url_path
+
+logger = logging.getLogger(__name__)
+
+
+def render_chart_image(
+    chart: Any,
+    dashboard_id: int,
+    active_data_mask: dict[str, Any],
+    user: Any,
+) -> bytes | None:
+    """
+    Render ``chart`` (as seen on ``dashboard_id``) to PNG bytes.
+
+    The chart is rendered through Explore in standalone mode with the live
+    dashboard filter state injected as ``extra_form_data`` — the same object the
+    data path merges into the query context — so the image and the data stay
+    consistent.
+
+    :param chart: The ``Slice`` to render
+    :param dashboard_id: The dashboard the chart is displayed on (for filter scope)
+    :param active_data_mask: Live dashboard filter state keyed by native filter id
+    :param user: The requesting user; the render runs with their permissions
+    :returns: PNG bytes, or ``None`` if the render failed (the caller skips and
+        notes the chart)
+    """
+    try:
+        filter_context = get_dashboard_filter_context(
+            dashboard_id=dashboard_id,
+            chart_id=chart.id,
+            active_data_mask=active_data_mask,
+        )
+
+        # Start from the chart's saved form data and force the slice id, then
+        # layer the live filters on top so the render matches the data path.
+        form_data: dict[str, Any] = json.loads(chart.params or "{}")
+        form_data["slice_id"] = chart.id
+        if filter_context.extra_form_data:
+            form_data["extra_form_data"] = filter_context.extra_form_data
+
+        url = get_url_path(
+            "ExploreView.root",
+            form_data=json.dumps(form_data),
+        )
+
+        window_size = current_app.config["WEBDRIVER_WINDOW"]["slice"]
+        screenshot = ChartScreenshot(
+            url,
+            chart.digest,
+            window_size=window_size,
+            thumb_size=window_size,
+        )
+        return screenshot.get_screenshot(user=user)
+    except SoftTimeLimitExceeded:
+        # A soft timeout aborts the whole export; don't let the broad handler
+        # below turn it into a ``None`` (a per-chart "could not render") result.
+        raise
+    except Exception:  # pylint: disable=broad-except
+        logger.exception(
+            "Failed to render image for chart %s in dashboard %s",
+            getattr(chart, "id", "?"),
+            dashboard_id,
+        )
+        return None
diff --git a/superset/dashboards/schemas.py b/superset/dashboards/schemas.py
index d0ffd31b285..b6e60f6224e 100644
--- a/superset/dashboards/schemas.py
+++ b/superset/dashboards/schemas.py
@@ -18,7 +18,7 @@ import re
 from typing import Any, Mapping, Union
 
 from marshmallow import fields, post_dump, post_load, pre_load, Schema
-from marshmallow.validate import Length, ValidationError
+from marshmallow.validate import Length, OneOf, ValidationError
 
 from superset import security_manager
 from superset.subjects.schemas import SubjectResponseSchema
@@ -628,3 +628,30 @@ class CacheScreenshotSchema(Schema):
         fields.List(fields.Str(), validate=lambda x: len(x) == 2), required=False
     )
     permalinkKey = fields.Str(required=False)  # noqa: N815
+
+
+class DashboardExportXlsxPostSchema(Schema):
+    active_data_mask = fields.Dict(
+        keys=fields.Str(),
+        values=fields.Dict(),
+        load_default=dict,
+        metadata={
+            "description": "Live dashboard filter state keyed by native filter id, "
+            "each carrying an extraFormData object."
+        },
+    )
+    mode = fields.String(
+        load_default="data",
+        validate=OneOf(["data", "images"]),
+        metadata={
+            "description": "Export mode: 'data' streams each chart's tabular result "
+            "(default); 'images' embeds non-table charts as rendered images and "
+            "keeps table charts tabular."
+        },
+    )
+
+
+class DashboardExportXlsxResponseSchema(Schema):
+    job_id = fields.String(
+        metadata={"description": "Correlation id for the async export task"}
+    )
diff --git a/superset/tasks/export_dashboard_excel.py b/superset/tasks/export_dashboard_excel.py
new file mode 100644
index 00000000000..a5f6035e024
--- /dev/null
+++ b/superset/tasks/export_dashboard_excel.py
@@ -0,0 +1,361 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+Celery task that exports every chart on a dashboard to a single multi-sheet
+``.xlsx`` file, uploads it to S3, and emails the requesting user a pre-signed
+download link.
+
+In ``"data"`` mode the task re-runs each chart's saved query context under the
+requesting user, applies the live dashboard filter state, and streams the results
+row-by-row into a constant-memory workbook so large dashboards never load all
+data at once. In ``"images"`` mode non-table charts are instead rendered to
+images (through the same headless path as scheduled reports, reflecting the live
+filters) and embedded, while table-like charts stay tabular.
+"""
+
+from __future__ import annotations
+
+import logging
+import os
+import tempfile
+from datetime import datetime, timedelta, timezone
+from typing import Any
+
+from celery.exceptions import SoftTimeLimitExceeded
+from flask import current_app, g
+
+from superset import db, security_manager
+from superset.charts.data.dashboard_filter_context import (
+    apply_dashboard_filter_context,
+    get_dashboard_filter_context,
+)
+from superset.charts.schemas import ChartDataQueryContextSchema
+from superset.commands.chart.data.get_data_command import ChartDataCommand
+from superset.commands.distributed_lock.release import ReleaseDistributedLock
+from superset.common.chart_data import ChartDataResultFormat, ChartDataResultType
+from superset.dashboards.excel_export import email
+from superset.dashboards.excel_export.layout import get_charts_in_layout_order
+from superset.dashboards.excel_export.screenshot import render_chart_image
+from superset.extensions import celery_app
+from superset.utils import json, s3
+from superset.utils.core import override_user
+from superset.utils.excel_streaming import StreamingXlsxWriter
+
+logger = logging.getLogger(__name__)
+
+# Export modes: "data" streams every chart's tabular result (the default,
+# unchanged behavior); "images" embeds non-table charts as rendered images and
+# keeps only table-like charts tabular.
+EXPORT_MODE_DATA = "data"
+EXPORT_MODE_IMAGES = "images"
+
+# Viz types kept as tabular data in image mode; everything else is rendered as an
+# image. Operators can override the set via ``EXCEL_EXPORT_TABLE_VIZ_TYPES``.
+TABLE_VIZ_TYPES = {"table", "pivot_table_v2", "pivot_table"}
+
+EXPORT_SOFT_TIME_LIMIT = 600
+EXPORT_HARD_TIME_LIMIT = 660
+
+# Namespace + TTL for the per-user+dashboard in-flight lock the API acquires
+# before enqueue and this task releases when it settles. The lock uses the
+# shared, atomic DistributedLock backend (Redis when configured, the metadata
+# DB otherwise) so it actually synchronizes across the web server and workers —
+# unlike a plain cache, which is a no-op under the default ``NullCache``.
+# The TTL outlives the hard time limit so a worker killed at that limit (which
+# skips the ``finally`` release) cannot hold the lock forever; the release in
+# ``finally`` is the fast path that frees it as soon as the task settles.
+EXPORT_LOCK_NAMESPACE = "excel_export"
+EXPORT_LOCK_TTL_SECONDS = EXPORT_HARD_TIME_LIMIT + 60
+
+
+def export_lock_params(user_id: int, dashboard_id: int) -> dict[str, int]:
+    """Key parameters identifying the per-user+dashboard in-flight lock."""
+    return {"user_id": user_id, "dashboard_id": dashboard_id}
+
+
+class _ChartSkippedError(Exception):
+    """Signals a chart that could not be exported and should be listed as skipped."""
+
+
+def _chart_label(chart: Any) -> str:
+    """Human-readable label for a chart in the skipped-charts list."""
+    return f"{chart.id} - {chart.slice_name or ''}".strip()
+
+
+def _record_to_row(record: dict[str, Any], colnames: list[str]) -> list[Any]:
+    return [record.get(col) for col in colnames]
+
+
+def _table_viz_types() -> set[str]:
+    """Viz types kept tabular in image mode (config override or built-in default)."""
+    return current_app.config.get("EXCEL_EXPORT_TABLE_VIZ_TYPES") or TABLE_VIZ_TYPES
+
+
+def _renders_as_image(chart: Any, mode: str) -> bool:
+    """Whether this chart is embedded as an image rather than streamed as data."""
+    return mode == EXPORT_MODE_IMAGES and chart.viz_type not in _table_viz_types()
+
+
+def _write_chart_image_sheet(
+    writer: StreamingXlsxWriter,
+    chart: Any,
+    dashboard_id: int,
+    active_data_mask: dict[str, Any],
+    user: Any,
+) -> None:
+    """
+    Render a single chart to an image and embed it as its own sheet.
+
+    :raises _ChartSkippedError: if the chart could not be rendered
+    """
+    image = render_chart_image(chart, dashboard_id, active_data_mask, user)
+    if image is None:
+        raise _ChartSkippedError
+    writer.add_image_sheet(_chart_label(chart), image)
+
+
+def _write_chart_sheets(
+    writer: StreamingXlsxWriter,
+    chart: Any,
+    dashboard_id: int,
+    active_data_mask: dict[str, Any],
+) -> None:
+    """
+    Run a single chart's query and stream its result(s) into the workbook.
+
+    Charts may yield more than one query (e.g. mixed-series charts); each becomes
+    its own sheet. Raises if the chart cannot be exported, so the caller can skip
+    it and note it in the email.
+    """
+    json_body = json.loads(chart.query_context)
+    # Override any stale saved values: we always want full JSON results.
+    json_body["result_format"] = ChartDataResultFormat.JSON
+    json_body["result_type"] = ChartDataResultType.FULL
+    json_body.pop("force", None)
+
+    filter_context = get_dashboard_filter_context(
+        dashboard_id=dashboard_id,
+        chart_id=chart.id,
+        active_data_mask=active_data_mask,
+    )
+    if filter_context.extra_form_data:
+        apply_dashboard_filter_context(json_body, filter_context.extra_form_data)
+
+    # Jinja macros resolve form data from g.form_data; expose the saved context.
+    g.form_data = json_body
+
+    query_context = ChartDataQueryContextSchema().load(json_body)
+    command = ChartDataCommand(query_context)
+    command.validate()
+    result = command.run()
+
+    for index, query in enumerate(result["queries"]):
+        colnames = query.get("colnames") or []
+        data = query.get("data") or []
+        if index == 0:
+            name = f"{chart.id} - {chart.slice_name or ''}"
+        else:
+            name = f"{chart.id}.{index} - {chart.slice_name or ''}"
+        writer.add_sheet(
+            name,
+            colnames,
+            (_record_to_row(record, colnames) for record in data),
+        )
+
+
+def _build_workbook(
+    path: str,
+    dashboard: Any,
+    active_data_mask: dict[str, Any],
+    job_id: str,
+    mode: str,
+    user: Any,
+) -> dict[str, list[str]]:
+    """Build the workbook on disk.
+
+    Return the charts that could not be exported, grouped by the reason they
+    were omitted (see the ``email.ERROR_*`` reason keys), so the notification
+    can explain each group separately.
+    """
+    errored: dict[str, list[str]] = {}
+    writer = StreamingXlsxWriter(path)
+    try:
+        for chart in get_charts_in_layout_order(dashboard):
+            label = _chart_label(chart)
+            as_image = _renders_as_image(chart, mode)
+            # Image charts render from their saved params and don't need a query
+            # context; data (and table) charts still do.
+            if not as_image and not chart.query_context:
+                errored.setdefault(email.ERROR_NO_QUERY_CONTEXT, []).append(label)
+                continue
+            try:
+                if as_image:
+                    _write_chart_image_sheet(
+                        writer, chart, dashboard.id, active_data_mask, user
+                    )
+                else:
+                    _write_chart_sheets(writer, chart, dashboard.id, active_data_mask)
+            except SoftTimeLimitExceeded:
+                # A soft timeout is a task-level signal, not a per-chart failure:
+                # let it propagate so the outer handler emails a failure and runs
+                # cleanup, rather than continuing until the hard limit kills the
+                # worker (which would skip cleanup, leak temp files, and hold the
+                # in-flight lock until its TTL). ``except Exception`` below would
+                # otherwise swallow it, since it subclasses ``Exception``.
+                raise
+            except _ChartSkippedError:
+                logger.warning(
+                    "Skipping chart %s in dashboard export %s (could not render)",
+                    chart.id,
+                    job_id,
+                )
+                errored.setdefault(email.ERROR_GENERAL, []).append(label)
+            except Exception:  # pylint: disable=broad-except
+                logger.exception(
+                    "Skipping chart %s in dashboard export %s", chart.id, job_id
+                )
+                errored.setdefault(email.ERROR_GENERAL, []).append(label)
+
+        if writer.sheet_count == 0:
+            flat = [label for labels in errored.values() for label in labels]
+            writer.add_summary_sheet(
+                "Export Summary",
+                ["No chart data could be exported.", *flat],
+            )
+    finally:
+        writer.close()
+    return errored
+
+
+def _send_failure_email(
+    user: Any, dashboard_title: str, requested_at: datetime
+) -> None:
+    if not (user and getattr(user, "email", None)):
+        return
+    try:
+        email.send_export_email(
+            user.email,
+            email.build_subject(dashboard_title, success=False),
+            email.build_failure_email(dashboard_title, requested_at),
+        )
+    except Exception:  # pylint: disable=broad-except
+        logger.exception("Failed to send export failure email")
+
+
+@celery_app.task(
+    name="export_dashboard_excel",
+    bind=True,
+    soft_time_limit=EXPORT_SOFT_TIME_LIMIT,
+    time_limit=EXPORT_HARD_TIME_LIMIT,
+    max_retries=0,
+)
+def export_dashboard_excel(
+    self: Any,  # pylint: disable=unused-argument
+    dashboard_id: int,
+    user_id: int,
+    active_data_mask: dict[str, Any],
+    job_id: str,
+    mode: str = EXPORT_MODE_DATA,
+) -> None:
+    """
+    Export a dashboard's charts to an ``.xlsx`` and email a download link.
+
+    :param dashboard_id: The dashboard to export
+    :param user_id: The requesting user (the task runs with their permissions)
+    :param active_data_mask: Live dashboard filter state keyed by native filter id
+    :param job_id: Correlation id, also the Celery task id and S3 object name
+    :param mode: ``"data"`` streams every chart's tabular result; ``"images"``
+        embeds non-table charts as rendered images and keeps tables tabular
+    """
+    # pylint: disable=import-outside-toplevel
+    from superset.models.dashboard import Dashboard
+
+    requested_at = datetime.now(tz=timezone.utc)
+    user = security_manager.get_user_by_id(user_id)
+    dashboard_title = ""
+    tmp_path: str | None = None
+
+    try:
+        with override_user(user, force=False):
+            dashboard = (
+                db.session.query(Dashboard).filter_by(id=dashboard_id).one_or_none()
+            )
+            if dashboard is None:
+                raise ValueError(f"Dashboard {dashboard_id} not found")
+            dashboard_title = dashboard.dashboard_title or f"Dashboard {dashboard_id}"
+
+            file_descriptor, tmp_path = tempfile.mkstemp(
+                suffix=".xlsx", prefix=f"dash-export-{job_id}-"
+            )
+            os.close(file_descriptor)
+
+            errored = _build_workbook(
+                tmp_path, dashboard, active_data_mask, job_id, mode, user
+            )
+
+            bucket = current_app.config["EXCEL_EXPORT_S3_BUCKET"]
+            key = (
+                f"{current_app.config['EXCEL_EXPORT_S3_KEY_PREFIX']}"
+                f"{dashboard_id}/{job_id}.xlsx"
+            )
+            ttl = current_app.config["EXCEL_EXPORT_LINK_TTL_SECONDS"]
+
+            s3.upload_file_to_s3(tmp_path, bucket, key)
+            download_url = s3.generate_presigned_url(bucket, key, ttl)
+            expires_at = datetime.now(tz=timezone.utc) + timedelta(seconds=ttl)
+
+            if user and getattr(user, "email", None):
+                try:
+                    email.send_export_email(
+                        user.email,
+                        email.build_subject(dashboard_title, success=True),
+                        email.build_success_email(
+                            dashboard_title=dashboard_title,
+                            download_url=download_url,
+                            requested_at=requested_at,
+                            expires_at=expires_at,
+                            ttl_seconds=ttl,
+                            errored=errored,
+                        ),
+                    )
+                except Exception:  # pylint: disable=broad-except
+                    # The file is already in S3; a send failure should not trigger
+                    # a misleading failure email.
+                    logger.exception("Failed to send export success email")
+    except SoftTimeLimitExceeded:
+        logger.warning("Dashboard excel export %s timed out", job_id)
+        _send_failure_email(user, dashboard_title, requested_at)
+        raise
+    except Exception:
+        logger.exception("Dashboard excel export %s failed", job_id)
+        _send_failure_email(user, dashboard_title, requested_at)
+        raise
+    finally:
+        try:
+            ReleaseDistributedLock(
+                EXPORT_LOCK_NAMESPACE,
+                export_lock_params(user_id, dashboard_id),
+            ).run()
+        except Exception:  # pylint: disable=broad-except
+            # Best-effort: the lock's TTL is the backstop if this fails.
+            logger.exception(
+                "Failed to release in-flight export lock for user %s dashboard %s",
+                user_id,
+                dashboard_id,
+            )
+        if tmp_path and os.path.exists(tmp_path):
+            os.remove(tmp_path)
diff --git a/superset/utils/excel.py b/superset/utils/excel.py
index 2af29e79a48..1a252de8fe5 100644
--- a/superset/utils/excel.py
+++ b/superset/utils/excel.py
@@ -41,18 +41,21 @@ NEUTRAL_DOCUMENT_PROPERTIES: dict[str, Any] = {
     "created": NEUTRAL_TIMESTAMP,
 }
 
+# Leading characters that turn a cell into a formula in spreadsheet apps. Shared
+# with the streaming writer (superset.utils.excel_streaming) so both export paths
+# guard against the same formula-injection vectors.
+FORMULA_PREFIXES = {"=", "+", "-", "@"}
+
 
 def quote_formulas(df: pd.DataFrame) -> pd.DataFrame:
     """
     Make sure to quote any formulas for security reasons.
     """
-    formula_prefixes = {"=", "+", "-", "@"}
-
     for col in df.select_dtypes(include="object").columns:
         df[col] = df[col].apply(
             lambda x: (
                 f"'{x}"
-                if isinstance(x, str) and len(x) and x[0] in formula_prefixes
+                if isinstance(x, str) and len(x) and x[0] in FORMULA_PREFIXES
                 else x
             )
         )
diff --git a/superset/utils/excel_streaming.py b/superset/utils/excel_streaming.py
new file mode 100644
index 00000000000..c9d93403ea1
--- /dev/null
+++ b/superset/utils/excel_streaming.py
@@ -0,0 +1,250 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+Streaming XLSX writer for multi-sheet dashboard exports.
+
+Unlike :mod:`superset.utils.excel`, which builds an in-memory DataFrame per
+sheet and hands the whole thing to ``xlsxwriter`` at once, this writer opens the
+workbook in ``constant_memory`` mode and writes rows one at a time, so
+``xlsxwriter`` keeps at most one row per sheet buffered on the writer side. The
+source records may still be materialized upstream (e.g. by the chart query
+response); this bounds only the writer's own footprint, not the caller's.
+"""
+
+from __future__ import annotations
+
+import math
+import numbers
+import re
+from collections.abc import Iterable, Sequence
+from datetime import date, datetime
+from decimal import Decimal
+from io import BytesIO
+from typing import Any
+
+import xlsxwriter
+
+from superset.utils.excel import FORMULA_PREFIXES, NEUTRAL_DOCUMENT_PROPERTIES
+
+# Excel limits a sheet name to 31 characters and forbids these characters.
+MAX_SHEET_NAME_LEN = 31
+_INVALID_SHEET_CHARS_RE = re.compile(r"[\[\]:*?/\\]")
+# Excel reserves the sheet name "History" (case-insensitive).
+_RESERVED_SHEET_NAME = "history"
+
+# A worksheet holds at most 1,048,576 rows; one is reserved for the header.
+MAX_DATA_ROWS_PER_SHEET = 1_048_576 - 1
+
+# Excel cannot represent integers beyond 10**15 without precision loss.
+_MAX_EXCEL_INT = 10**15
+
+
+def _quote_if_formula(text: str) -> str:
+    """
+    Prefix formula-like text with an apostrophe so spreadsheet apps treat it as
+    literal text (defense against formula injection).
+
+    Leading whitespace is ignored when detecting a formula, because spreadsheet
+    apps still evaluate a cell whose formula prefix is preceded by spaces or
+    tabs (e.g. ``" =cmd"`` or ``"\\t=cmd"``).
+    """
+    stripped = text.lstrip()
+    return f"'{text}" if stripped and stripped[0] in FORMULA_PREFIXES else text
+
+
+def _coerce_float_cell(value: Any) -> Any:
+    """
+    Convert a ``Decimal``/real value to something ``xlsxwriter`` accepts.
+
+    ``float()`` on a non-finite ``Decimal`` ("NaN"/"Infinity") yields a value
+    xlsxwriter rejects, and an over-large value can raise ``OverflowError``;
+    blank the former and stringify the latter, and stringify magnitudes Excel
+    cannot represent precisely.
+    """
+    try:
+        number = float(value)
+    except (OverflowError, ValueError):
+        return str(value)
+    if not math.isfinite(number):
+        return ""
+    return str(number) if abs(number) > _MAX_EXCEL_INT else number
+
+
+def sanitize_sheet_name(raw: str, used: set[str]) -> str:
+    """
+    Produce a valid, unique Excel sheet name from ``raw``.
+
+    Replaces forbidden characters, strips surrounding apostrophes/whitespace,
+    avoids the reserved name "History", truncates to 31 characters, and
+    disambiguates case-insensitive collisions with ``~2``/``~3`` suffixes.
+    The chosen name (lower-cased) is added to ``used``.
+
+    :param raw: The desired sheet name (e.g. ``"42 - Sales by Region"``)
+    :param used: Lower-cased names already taken; mutated with the result
+    :returns: A sanitized, unique sheet name no longer than 31 characters
+    """
+    name = _INVALID_SHEET_CHARS_RE.sub("_", raw or "")
+    name = name.strip().strip("'").strip()
+    if not name:
+        name = "Sheet"
+    if name.lower() == _RESERVED_SHEET_NAME:
+        name = f"{name}_"
+    name = name[:MAX_SHEET_NAME_LEN]
+
+    if name.lower() not in used:
+        used.add(name.lower())
+        return name
+
+    suffix = 2
+    while True:
+        marker = f"~{suffix}"
+        candidate = name[: MAX_SHEET_NAME_LEN - len(marker)] + marker
+        if candidate.lower() not in used:
+            used.add(candidate.lower())
+            return candidate
+        suffix += 1
+
+
+def _sanitize_cell(value: Any) -> Any:
+    """
+    Coerce a single cell value into something safe for ``xlsxwriter``.
+
+    Quotes formula-like strings (defense against formula injection), stringifies
+    integers/floats Excel cannot represent precisely, renders temporal values as
+    ISO strings (timezones are not natively supported), and blanks out ``None``
+    and non-finite floats.
+    """
+    if value is None:
+        return ""
+    # bool is a subclass of int; preserve it before the numeric branches.
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        return _quote_if_formula(value)
+    if isinstance(value, (datetime, date)):
+        return value.isoformat()
+    if isinstance(value, Decimal):
+        return _coerce_float_cell(value)
+    if isinstance(value, numbers.Integral):
+        number = int(value)
+        return str(number) if abs(number) > _MAX_EXCEL_INT else number
+    if isinstance(value, numbers.Real):
+        return _coerce_float_cell(value)
+    # Anything else (lists, dicts, custom objects) is stringified, still guarding
+    # against formula injection on the resulting text.
+    return _quote_if_formula(str(value))
+
+
+class StreamingXlsxWriter:
+    """
+    A thin wrapper over ``xlsxwriter`` in constant-memory mode that writes one
+    sheet per chart, row by row.
+
+    Sheet names are sanitized and de-duplicated, cell values are sanitized for
+    safety/compatibility, and per-sheet row counts are capped at Excel's limit.
+    Always call :meth:`close` (e.g. in a ``finally`` block) to finalize the file.
+    """
+
+    def __init__(self, path: str) -> None:
+        self._workbook = xlsxwriter.Workbook(path, {"constant_memory": True})
+        # Reset document properties so the file carries no identifying details.
+        self._workbook.set_properties(NEUTRAL_DOCUMENT_PROPERTIES)
+        self._used_sheet_names: set[str] = set()
+        self.sheet_count = 0
+
+    def add_sheet(
+        self,
+        name: str,
+        columns: Sequence[Any],
+        rows: Iterable[Sequence[Any]],
+    ) -> int:
+        """
+        Write a header row followed by data rows into a new sheet.
+
+        :param name: Desired sheet name (sanitized/de-duplicated automatically)
+        :param columns: Column headers
+        :param rows: Iterable of row sequences, streamed one at a time
+        :returns: The number of data rows actually written (capped just below
+            Excel's per-sheet limit; when the data is larger a final notice row
+            is appended and the dropped rows are not counted)
+        """
+        sheet_name = sanitize_sheet_name(name, self._used_sheet_names)
+        worksheet = self._workbook.add_worksheet(sheet_name)
+        worksheet.write_row(0, 0, [_sanitize_cell(col) for col in columns])
+
+        # Reserve the final row for a truncation notice, so when the data
+        # exceeds the sheet's capacity the user can see rows were dropped
+        # instead of silently losing them.
+        row_cap = MAX_DATA_ROWS_PER_SHEET - 1
+        written = 0
+        truncated = False
+        for row in rows:
+            if written >= row_cap:
+                truncated = True
+                break
+            worksheet.write_row(written + 1, 0, [_sanitize_cell(cell) for cell in row])
+            written += 1
+
+        if truncated:
+            worksheet.write_string(
+                written + 1,
+                0,
+                f"[Truncated: only first {written:,} rows exported]",
+            )
+
+        self.sheet_count += 1
+        return written
+
+    def add_image_sheet(self, name: str, image_bytes: bytes) -> None:
+        """
+        Write a single sheet holding a rendered chart image.
+
+        The image is embedded top-left; ``xlsxwriter`` buffers image data and
+        writes it at :meth:`close`, so this composes with ``constant_memory``
+        mode just like the row-streaming sheets.
+
+        :param name: Desired sheet name (sanitized/de-duplicated automatically)
+        :param image_bytes: PNG bytes to embed
+        """
+        sheet_name = sanitize_sheet_name(name, self._used_sheet_names)
+        worksheet = self._workbook.add_worksheet(sheet_name)
+        worksheet.insert_image(
+            0,
+            0,
+            f"{sheet_name}.png",
+            {"image_data": BytesIO(image_bytes)},
+        )
+        self.sheet_count += 1
+
+    def add_summary_sheet(self, name: str, lines: Sequence[str]) -> None:
+        """
+        Write a single-column informational sheet (e.g. a list of skipped charts).
+
+        Lines are written as string cells, so formula-like text is never executed.
+        """
+        sheet_name = sanitize_sheet_name(name, self._used_sheet_names)
+        worksheet = self._workbook.add_worksheet(sheet_name)
+        for index, line in enumerate(lines):
+            worksheet.write_string(index, 0, str(line))
+        self.sheet_count += 1
+
+    def close(self) -> None:
+        """Finalize and write the workbook to disk."""
+        if self.sheet_count == 0:
+            # Excel requires at least one worksheet for a valid file.
+            self._workbook.add_worksheet("Export")
+        self._workbook.close()
diff --git a/superset/utils/s3.py b/superset/utils/s3.py
new file mode 100644
index 00000000000..1e68b980964
--- /dev/null
+++ b/superset/utils/s3.py
@@ -0,0 +1,83 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+Minimal S3 helpers for uploading export artifacts and minting pre-signed URLs.
+
+Credentials and region come from the standard boto3 resolution chain (env vars,
+shared config, instance role). Operators can override client construction via
+the ``EXCEL_EXPORT_S3_CLIENT_KWARGS`` config (e.g. ``region_name`` or an
+``endpoint_url`` for S3-compatible stores such as MinIO/LocalStack).
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+from flask import current_app
+
+logger = logging.getLogger(__name__)
+
+
+def _get_s3_client() -> Any:
+    """Build an S3 client using operator-provided client kwargs (if any)."""
+    # boto3 is imported lazily so that importing this module (which happens at
+    # app startup via the dashboard API) does not require boto3 to be installed.
+    # The dependency is only needed when an export actually runs; if it is
+    # missing, surface an actionable install hint rather than a bare ImportError.
+    try:
+        import boto3  # pylint: disable=import-outside-toplevel
+    except ImportError as ex:
+        raise ImportError(
+            "boto3 is required for dashboard Excel export but is not installed. "
+            "Install it with `pip install apache-superset[excel-export]`."
+        ) from ex
+
+    client_kwargs: dict[str, Any] = current_app.config.get(
+        "EXCEL_EXPORT_S3_CLIENT_KWARGS", {}
+    )
+    return boto3.client("s3", **client_kwargs)
+
+
+def upload_file_to_s3(local_path: str, bucket: str, key: str) -> None:
+    """
+    Upload a local file to S3.
+
+    ``boto3``'s ``upload_file`` automatically uses a managed multipart transfer
+    for large files, so no manual chunking is required.
+
+    :param local_path: Path to the file on local disk
+    :param bucket: Destination S3 bucket
+    :param key: Destination S3 object key
+    """
+    _get_s3_client().upload_file(local_path, bucket, key)
+
+
+def generate_presigned_url(bucket: str, key: str, expires_in: int) -> str:
+    """
+    Generate a time-limited pre-signed URL for downloading an S3 object.
+
+    :param bucket: The S3 bucket
+    :param key: The S3 object key
+    :param expires_in: URL lifetime in seconds
+    :returns: A pre-signed ``get_object`` URL
+    """
+    return _get_s3_client().generate_presigned_url(
+        "get_object",
+        Params={"Bucket": bucket, "Key": key},
+        ExpiresIn=expires_in,
+    )
diff --git a/tests/integration_tests/dashboards/api_tests.py b/tests/integration_tests/dashboards/api_tests.py
index a87bb797d0c..a25af9fcd48 100644
--- a/tests/integration_tests/dashboards/api_tests.py
+++ b/tests/integration_tests/dashboards/api_tests.py
@@ -31,6 +31,7 @@ import yaml
 from freezegun import freeze_time
 from sqlalchemy import and_
 from superset import db, security_manager  # noqa: F401
+from superset.exceptions import LockAlreadyHeldException
 from superset.models.dashboard import Dashboard
 from superset.models.core import FavStar, FavStarClassName
 from superset.reports.models import ReportSchedule, ReportScheduleType
@@ -42,6 +43,7 @@ from superset.utils.core import backend, override_user
 from superset.utils.screenshots import ScreenshotCachePayload
 from superset.utils import json
 
+from tests.conftest import with_config
 from tests.integration_tests.base_api_tests import ApiEditorsTestCaseMixin
 from tests.integration_tests.base_tests import (
     subjects_from_users,
@@ -3262,6 +3264,173 @@ class TestDashboardApi(ApiEditorsTestCaseMixin, InsertChartMixin, SupersetTestCa
         response = json.loads(rv.data.decode("utf-8"))
         assert response["count"] > 0
 
+    def test_export_xlsx_501_when_bucket_unset(self):
+        """Dashboard API: export_xlsx returns 501 when the S3 bucket is unset."""
+        admin = self.get_user("admin")
+        dashboard = self.insert_dashboard("xlsx-501", None, [admin.id])
+        self.login(ADMIN_USERNAME)
+        try:
+            rv = self.client.post(f"api/v1/dashboard/{dashboard.id}/export_xlsx/")
+            assert rv.status_code == 501
+        finally:
+            db.session.delete(dashboard)
+            db.session.commit()
+
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_404_for_missing_dashboard(self, mock_task):
+        """Dashboard API: export_xlsx returns 404 for an unknown dashboard."""
+        self.login(ADMIN_USERNAME)
+        rv = self.client.post("api/v1/dashboard/99999999/export_xlsx/")
+        assert rv.status_code == 404
+        mock_task.apply_async.assert_not_called()
+
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_400_for_empty_dashboard(self, mock_task):
+        """Dashboard API: export_xlsx returns 400 for a dashboard with no charts."""
+        admin = self.get_user("admin")
+        dashboard = self.insert_dashboard("xlsx-empty", None, [admin.id])
+        self.login(ADMIN_USERNAME)
+        try:
+            rv = self.client.post(f"api/v1/dashboard/{dashboard.id}/export_xlsx/")
+            assert rv.status_code == 400
+            mock_task.apply_async.assert_not_called()
+        finally:
+            db.session.delete(dashboard)
+            db.session.commit()
+
+    @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.AcquireDistributedLock")
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_202_enqueues_task(self, mock_task, mock_acquire):
+        """Dashboard API: export_xlsx enqueues the task and returns 202 + job_id."""
+        self.login(ADMIN_USERNAME)
+        dashboard = db.session.query(Dashboard).filter_by(slug="world_health").first()
+        rv = self.client.post(
+            f"api/v1/dashboard/{dashboard.id}/export_xlsx/",
+            json={"active_data_mask": {}},
+        )
+        assert rv.status_code == 202
+        body = json.loads(rv.data.decode("utf-8"))
+        job_id = body["job_id"]
+        assert job_id
+        # The in-flight lock is acquired before the task is enqueued.
+        mock_acquire.return_value.run.assert_called_once()
+        mock_task.apply_async.assert_called_once()
+        _, kwargs = mock_task.apply_async.call_args
+        assert kwargs["task_id"] == job_id
+        assert kwargs["kwargs"]["dashboard_id"] == dashboard.id
+
+    @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.AcquireDistributedLock")
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_202_when_export_already_in_progress(
+        self, mock_task, mock_acquire
+    ):
+        """Dashboard API: export_xlsx does not enqueue a second concurrent export."""
+        # An in-flight lock is already held for this user+dashboard.
+        mock_acquire.return_value.run.side_effect = LockAlreadyHeldException("held")
+        self.login(ADMIN_USERNAME)
+        dashboard = db.session.query(Dashboard).filter_by(slug="world_health").first()
+        rv = self.client.post(
+            f"api/v1/dashboard/{dashboard.id}/export_xlsx/",
+            json={"active_data_mask": {}},
+        )
+        assert rv.status_code == 202
+        assert "already in progress" in rv.data.decode("utf-8")
+        mock_task.apply_async.assert_not_called()
+
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_404_for_inaccessible_dashboard(self, mock_task):
+        """Dashboard API: export_xlsx returns 404 for a dashboard the user can't see."""
+        admin = self.get_user("admin")
+        dashboard = self.insert_dashboard(
+            "xlsx-private", None, [admin.id], published=False
+        )
+        self.login(GAMMA_USERNAME)
+        try:
+            rv = self.client.post(f"api/v1/dashboard/{dashboard.id}/export_xlsx/")
+            assert rv.status_code == 404
+            mock_task.apply_async.assert_not_called()
+        finally:
+            db.session.delete(dashboard)
+            db.session.commit()
+
+    @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.AcquireDistributedLock")
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    @patch("superset.dashboards.api.security_manager.raise_for_access")
+    def test_export_xlsx_admitted_with_can_export_only(
+        self, mock_raise, mock_task, mock_acquire
+    ):
+        """Dashboard API: export_xlsx is gated on ``can_export``, not a distinct
+        ``can_export_xlsx``. Gamma holds dashboard ``can_export`` by default (and
+        the frontend shows the menu item on that basis), so a Gamma user must be
+        admitted (202) rather than rejected by ``@protect()`` (403)."""
+        gamma_user = security_manager.find_user(username=GAMMA_USERNAME)
+        slice_ = db.session.query(Slice).first()
+        # Clone Gamma (so the login password is valid); Gamma already carries
+        # dashboard ``can_export``.
+        with self.temporary_user(gamma_user, login=True) as user:
+            dashboard = self.insert_dashboard(
+                "xlsx-can-export", None, [user.id], slices=[slice_], published=True
+            )
+            try:
+                rv = self.client.post(
+                    f"api/v1/dashboard/{dashboard.id}/export_xlsx/",
+                    json={"active_data_mask": {}},
+                )
+                assert rv.status_code == 202
+                mock_task.apply_async.assert_called_once()
+            finally:
+                db.session.delete(dashboard)
+                db.session.commit()
+
+    @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_images_404_when_screenshot_flags_off(self, mock_task):
+        """Dashboard API: ``mode=images`` is rejected with 404 when the webdriver
+        screenshot flags are disabled (the same signal the UI gates the option on),
+        and no task is enqueued."""
+        self.login(ADMIN_USERNAME)
+        dashboard = db.session.query(Dashboard).filter_by(slug="world_health").first()
+        rv = self.client.post(
+            f"api/v1/dashboard/{dashboard.id}/export_xlsx/",
+            json={"active_data_mask": {}, "mode": "images"},
+        )
+        assert rv.status_code == 404
+        mock_task.apply_async.assert_not_called()
+
+    @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
+    @with_config({"EXCEL_EXPORT_S3_BUCKET": "exports"})
+    @with_feature_flags(
+        ENABLE_DASHBOARD_SCREENSHOT_ENDPOINTS=True,
+        ENABLE_DASHBOARD_DOWNLOAD_WEBDRIVER_SCREENSHOT=True,
+    )
+    @patch("superset.dashboards.api.AcquireDistributedLock")
+    @patch("superset.dashboards.api.export_dashboard_excel")
+    def test_export_xlsx_images_202_when_screenshot_flags_on(
+        self, mock_task, mock_acquire
+    ):
+        """Dashboard API: ``mode=images`` is accepted (202) when both webdriver
+        screenshot flags are enabled."""
+        self.login(ADMIN_USERNAME)
+        dashboard = db.session.query(Dashboard).filter_by(slug="world_health").first()
+        rv = self.client.post(
+            f"api/v1/dashboard/{dashboard.id}/export_xlsx/",
+            json={"active_data_mask": {}, "mode": "images"},
+        )
+        assert rv.status_code == 202
+        mock_task.apply_async.assert_called_once()
+        _, kwargs = mock_task.apply_async.call_args
+        assert kwargs["kwargs"]["mode"] == "images"
+
     @pytest.mark.usefixtures("load_world_bank_dashboard_with_slices")
     def test_embedded_dashboards(self):
         self.login(ADMIN_USERNAME)
diff --git a/tests/unit_tests/charts/test_dashboard_filter_context.py b/tests/unit_tests/charts/test_dashboard_filter_context.py
index 3bc062330bf..f784eeb744a 100644
--- a/tests/unit_tests/charts/test_dashboard_filter_context.py
+++ b/tests/unit_tests/charts/test_dashboard_filter_context.py
@@ -563,6 +563,178 @@ def test_get_dashboard_filter_context_out_of_scope_filter_excluded(
     assert ctx.filters[0].id == "f1"
 
 
+def _build_dashboard_mock(
+    mock_db: MagicMock,
+    filter_config: list[dict[str, Any]],
+    chart_ids: list[int],
+) -> MagicMock:
+    """Wire a dashboard MagicMock with the given filters and chart ids."""
+    metadata = {"native_filter_configuration": filter_config}
+    dashboard = MagicMock()
+    dashboard.id = 1
+    dashboard.slices = [MagicMock(id=cid) for cid in chart_ids]
+    dashboard.json_metadata = json.dumps(metadata)
+    dashboard.position_json = json.dumps(SAMPLE_POSITION_JSON)
+    (
+        mock_db.session.query.return_value.filter_by.return_value.one_or_none.return_value
+    ) = dashboard
+    return dashboard
+
+
+@patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
+@patch("superset.charts.data.dashboard_filter_context.db")
+def test_active_data_mask_overrides_default(
+    mock_db: MagicMock,
+    mock_check_access: MagicMock,
+) -> None:
+    """An active filter value replaces the saved default."""
+    filter_config = [
+        _make_filter(
+            flt_id="f1",
+            name="Region",
+            scope_root=["ROOT_ID"],
+            default_value=["US"],
+            target_column="region",
+        ),
+    ]
+    _build_dashboard_mock(mock_db, filter_config, [10])
+
+    active_data_mask = {
+        "f1": {
+            "extraFormData": {
+                "filters": [{"col": "region", "op": "IN", "val": ["FR", "DE"]}]
+            }
+        }
+    }
+    ctx = get_dashboard_filter_context(
+        dashboard_id=1, chart_id=10, active_data_mask=active_data_mask
+    )
+
+    assert ctx.filters[0].status == DashboardFilterStatus.APPLIED
+    assert ctx.extra_form_data["filters"][0]["val"] == ["FR", "DE"]
+
+
+@patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
+@patch("superset.charts.data.dashboard_filter_context.db")
+def test_active_data_mask_empty_clears_default(
+    mock_db: MagicMock,
+    mock_check_access: MagicMock,
+) -> None:
+    """An empty active extraFormData clears the filter; the default is NOT used."""
+    filter_config = [
+        _make_filter(
+            flt_id="f1",
+            name="Region",
+            scope_root=["ROOT_ID"],
+            default_value=["US"],
+            target_column="region",
+        ),
+    ]
+    _build_dashboard_mock(mock_db, filter_config, [10])
+
+    active_data_mask: dict[str, Any] = {"f1": {"extraFormData": {}}}
+    ctx = get_dashboard_filter_context(
+        dashboard_id=1, chart_id=10, active_data_mask=active_data_mask
+    )
+
+    assert ctx.filters[0].status == DashboardFilterStatus.NOT_APPLIED
+    assert "filters" not in ctx.extra_form_data
+
+
+@patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
+@patch("superset.charts.data.dashboard_filter_context.db")
+def test_active_data_mask_absent_filter_falls_back_to_default(
+    mock_db: MagicMock,
+    mock_check_access: MagicMock,
+) -> None:
+    """A filter not present in the mask keeps its saved default."""
+    filter_config = [
+        _make_filter(
+            flt_id="f1",
+            name="Region",
+            scope_root=["ROOT_ID"],
+            default_value=["US"],
+            target_column="region",
+        ),
+    ]
+    _build_dashboard_mock(mock_db, filter_config, [10])
+
+    # Mask only references some other filter id
+    active_data_mask: dict[str, Any] = {"f2": {"extraFormData": {"filters": []}}}
+    ctx = get_dashboard_filter_context(
+        dashboard_id=1, chart_id=10, active_data_mask=active_data_mask
+    )
+
+    assert ctx.filters[0].status == DashboardFilterStatus.APPLIED
+    assert ctx.extra_form_data["filters"][0]["val"] == ["US"]
+
+
+@patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
+@patch("superset.charts.data.dashboard_filter_context.db")
+def test_active_data_mask_applies_despite_default_to_first_item(
+    mock_db: MagicMock,
+    mock_check_access: MagicMock,
+) -> None:
+    """
+    defaultToFirstItem filters cannot be resolved from saved config, but when the
+    frontend supplies a concrete active value it is applied.
+    """
+    filter_config = [
+        _make_filter(
+            flt_id="f1",
+            name="City",
+            scope_root=["ROOT_ID"],
+            default_to_first_item=True,
+            target_column="city",
+        ),
+    ]
+    _build_dashboard_mock(mock_db, filter_config, [10])
+
+    active_data_mask = {
+        "f1": {
+            "extraFormData": {"filters": [{"col": "city", "op": "IN", "val": ["NYC"]}]}
+        }
+    }
+    ctx = get_dashboard_filter_context(
+        dashboard_id=1, chart_id=10, active_data_mask=active_data_mask
+    )
+
+    assert ctx.filters[0].status == DashboardFilterStatus.APPLIED
+    assert ctx.extra_form_data["filters"][0]["val"] == ["NYC"]
+
+
+@patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
+@patch("superset.charts.data.dashboard_filter_context.db")
+def test_active_data_mask_out_of_scope_filter_still_excluded(
+    mock_db: MagicMock,
+    mock_check_access: MagicMock,
+) -> None:
+    """An active value for an out-of-scope filter does not leak into the chart."""
+    filter_config = [
+        _make_filter(
+            flt_id="f1",
+            name="Out-of-scope",
+            scope_root=["TABS-nonexistent"],
+            target_column="status",
+        ),
+    ]
+    _build_dashboard_mock(mock_db, filter_config, [10])
+
+    active_data_mask = {
+        "f1": {
+            "extraFormData": {
+                "filters": [{"col": "status", "op": "IN", "val": ["active"]}]
+            }
+        }
+    }
+    ctx = get_dashboard_filter_context(
+        dashboard_id=1, chart_id=10, active_data_mask=active_data_mask
+    )
+
+    assert ctx.filters == []
+    assert ctx.extra_form_data == {}
+
+
 @patch("superset.charts.data.dashboard_filter_context._check_dashboard_access")
 @patch("superset.charts.data.dashboard_filter_context.db")
 def test_get_dashboard_filter_context_chart_not_in_layout_receives_root_filters(
diff --git a/tests/unit_tests/dashboards/test_excel_export_email.py b/tests/unit_tests/dashboards/test_excel_export_email.py
new file mode 100644
index 00000000000..de9bc294842
--- /dev/null
+++ b/tests/unit_tests/dashboards/test_excel_export_email.py
@@ -0,0 +1,157 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+from datetime import datetime
+from unittest.mock import MagicMock, patch
+
+from superset.dashboards.excel_export import email
+
+REQUESTED = datetime(2026, 1, 1, 12, 0, 0)
+EXPIRES = datetime(2026, 1, 2, 12, 0, 0)
+
+
+def test_success_email_contains_link_and_expiry() -> None:
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://signed.example/file.xlsx?sig=abc",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=86400,
+        errored={},
+    )
+    assert "https://signed.example/file.xlsx?sig=abc" in html
+    assert "expires in 24 hours" in html
+    assert "2026-01-02 12:00:00 UTC" in html
+    assert "2026-01-01 12:00:00 UTC" in html
+    assert "<li>" not in html  # no skipped section
+
+
+def test_success_email_sub_hour_ttl_reports_minutes() -> None:
+    # A sub-hour TTL must not truncate to "0 hours"; it should report minutes.
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=900,
+        errored={},
+    )
+    assert "expires in 15 minutes" in html
+    assert "0 hours" not in html
+
+
+def test_success_email_mixed_ttl_reports_hours_and_minutes() -> None:
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=5400,
+        errored={},
+    )
+    assert "expires in 1 hour 30 minutes" in html
+
+
+def test_success_email_lists_charts_with_no_query_context() -> None:
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=86400,
+        errored={email.ERROR_NO_QUERY_CONTEXT: ["10 - Broken chart"]},
+    )
+    assert "no saved query context" in html
+    assert "<li>10 - Broken chart</li>" in html
+
+
+def test_success_email_groups_errors_by_reason() -> None:
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=86400,
+        errored={
+            email.ERROR_NO_QUERY_CONTEXT: ["10 - NoContext"],
+            email.ERROR_GENERAL: ["30 - Boom"],
+        },
+    )
+    # Each reason renders its own labelled section with the right chart.
+    assert "no saved query context" in html
+    assert "an error occurred" in html
+    assert "<li>10 - NoContext</li>" in html
+    assert "<li>30 - Boom</li>" in html
+
+
+def test_success_email_omits_empty_reason_groups() -> None:
+    html = email.build_success_email(
+        dashboard_title="Sales",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=86400,
+        errored={email.ERROR_GENERAL: ["30 - Boom"]},
+    )
+    assert "an error occurred" in html
+    assert "no saved query context" not in html
+    assert "<li>30 - Boom</li>" in html
+
+
+def test_success_email_escapes_title() -> None:
+    html = email.build_success_email(
+        dashboard_title="<script>alert(1)</script>",
+        download_url="https://x",
+        requested_at=REQUESTED,
+        expires_at=EXPIRES,
+        ttl_seconds=86400,
+        errored={},
+    )
+    assert "<script>" not in html
+    assert "&lt;script&gt;" in html
+
+
+def test_failure_email_body() -> None:
+    html = email.build_failure_email("Sales", REQUESTED)
+    assert "could not be completed" in html
+    assert "2026-01-01 12:00:00 UTC" in html
+
+
+@patch("superset.dashboards.excel_export.email.current_app")
+def test_build_subject(mock_app: MagicMock) -> None:
+    mock_app.config = {"EMAIL_REPORTS_SUBJECT_PREFIX": "[Report] "}
+    assert (
+        email.build_subject("Sales", success=True)
+        == "[Report] Your dashboard export is ready: Sales"
+    )
+    assert email.build_subject("Sales", success=False).startswith(
+        "[Report] Your dashboard export could not be completed"
+    )
+
+
+@patch("superset.dashboards.excel_export.email.send_email_smtp")
+@patch("superset.dashboards.excel_export.email.current_app")
+def test_send_export_email(mock_app: MagicMock, mock_send: MagicMock) -> None:
+    mock_app.config = {"SMTP_HOST": "localhost"}
+    email.send_export_email("user@example.com", "subj", "<html></html>")
+    mock_send.assert_called_once_with(
+        to="user@example.com",
+        subject="subj",
+        html_content="<html></html>",
+        config=mock_app.config,
+    )
diff --git a/tests/unit_tests/dashboards/test_excel_export_layout.py b/tests/unit_tests/dashboards/test_excel_export_layout.py
new file mode 100644
index 00000000000..c7fc346439d
--- /dev/null
+++ b/tests/unit_tests/dashboards/test_excel_export_layout.py
@@ -0,0 +1,105 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+from typing import Any
+from unittest.mock import MagicMock
+
+from superset.dashboards.excel_export.layout import get_charts_in_layout_order
+
+
+def _chart_node(node_id: str, chart_id: int) -> dict[str, Any]:
+    return {"id": node_id, "type": "CHART", "meta": {"chartId": chart_id}}
+
+
+def _dashboard(position: dict[str, Any], chart_ids: list[int]) -> MagicMock:
+    dashboard = MagicMock()
+    dashboard.position = position
+    dashboard.slices = [MagicMock(id=cid) for cid in chart_ids]
+    return dashboard
+
+
+def _ids(slices: list[Any]) -> list[int]:
+    return [slc.id for slc in slices]
+
+
+def test_grid_order() -> None:
+    position = {
+        "ROOT_ID": {"type": "ROOT", "children": ["GRID_ID"]},
+        "GRID_ID": {"type": "GRID", "children": ["ROW-1"]},
+        "ROW-1": {"type": "ROW", "children": ["CHART-a", "CHART-b"]},
+        "CHART-a": _chart_node("CHART-a", 1),
+        "CHART-b": _chart_node("CHART-b", 2),
+    }
+    dashboard = _dashboard(position, [2, 1])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [1, 2]
+
+
+def test_tab_nested_order() -> None:
+    position = {
+        "ROOT_ID": {"type": "ROOT", "children": ["GRID_ID"]},
+        "GRID_ID": {"type": "GRID", "children": ["TABS-1"]},
+        "TABS-1": {"type": "TABS", "children": ["TAB-1", "TAB-2"]},
+        "TAB-1": {"type": "TAB", "children": ["CHART-a"]},
+        "TAB-2": {"type": "TAB", "children": ["CHART-b"]},
+        "CHART-a": _chart_node("CHART-a", 10),
+        "CHART-b": _chart_node("CHART-b", 20),
+    }
+    dashboard = _dashboard(position, [20, 10])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [10, 20]
+
+
+def test_duplicate_chart_placement_exported_once() -> None:
+    position = {
+        "ROOT_ID": {"type": "ROOT", "children": ["GRID_ID"]},
+        "GRID_ID": {"type": "GRID", "children": ["ROW-1", "ROW-2"]},
+        "ROW-1": {"type": "ROW", "children": ["CHART-a"]},
+        "ROW-2": {"type": "ROW", "children": ["CHART-a-dup"]},
+        "CHART-a": _chart_node("CHART-a", 5),
+        "CHART-a-dup": _chart_node("CHART-a-dup", 5),
+    }
+    dashboard = _dashboard(position, [5])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [5]
+
+
+def test_orphan_charts_appended_by_id() -> None:
+    position = {
+        "ROOT_ID": {"type": "ROOT", "children": ["GRID_ID"]},
+        "GRID_ID": {"type": "GRID", "children": ["ROW-1"]},
+        "ROW-1": {"type": "ROW", "children": ["CHART-a"]},
+        "CHART-a": _chart_node("CHART-a", 7),
+    }
+    # Charts 3 and 9 are on the dashboard but not in the layout.
+    dashboard = _dashboard(position, [7, 9, 3])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [7, 3, 9]
+
+
+def test_stale_layout_chart_skipped() -> None:
+    position = {
+        "ROOT_ID": {"type": "ROOT", "children": ["GRID_ID"]},
+        "GRID_ID": {"type": "GRID", "children": ["ROW-1"]},
+        "ROW-1": {"type": "ROW", "children": ["CHART-a", "CHART-gone"]},
+        "CHART-a": _chart_node("CHART-a", 1),
+        "CHART-gone": _chart_node("CHART-gone", 999),  # not in dashboard.slices
+    }
+    dashboard = _dashboard(position, [1])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [1]
+
+
+def test_empty_position_returns_all_slices_sorted() -> None:
+    dashboard = _dashboard({}, [3, 1, 2])
+    assert _ids(get_charts_in_layout_order(dashboard)) == [1, 2, 3]
diff --git a/tests/unit_tests/dashboards/test_excel_export_screenshot.py b/tests/unit_tests/dashboards/test_excel_export_screenshot.py
new file mode 100644
index 00000000000..d4455a98792
--- /dev/null
+++ b/tests/unit_tests/dashboards/test_excel_export_screenshot.py
@@ -0,0 +1,164 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+from typing import Any
+from unittest.mock import MagicMock, patch
+
+import pytest
+from celery.exceptions import SoftTimeLimitExceeded
+
+from superset.charts.data.dashboard_filter_context import DashboardFilterContext
+from superset.dashboards.excel_export import screenshot as screenshot_module
+from superset.dashboards.excel_export.screenshot import render_chart_image
+from superset.utils import json
+
+MODULE = "superset.dashboards.excel_export.screenshot"
+
+
+def _chart(
+    chart_id: int = 42,
+    params: str = '{"viz_type": "pie"}',
+    digest: str = "abc123",
+) -> MagicMock:
+    chart = MagicMock()
+    chart.id = chart_id
+    chart.params = params
+    chart.digest = digest
+    return chart
+
+
+def _patched_app() -> Any:
+    app = MagicMock()
+    app.config = {"WEBDRIVER_WINDOW": {"slice": (3000, 1200)}}
+    return app
+
+
+def test_render_chart_image_builds_url_with_slice_id_and_filters() -> None:
+    chart = _chart()
+    active_mask = {"NATIVE_FILTER-1": {"extraFormData": {"filters": [{"col": "a"}]}}}
+    filter_context = DashboardFilterContext(
+        extra_form_data={"filters": [{"col": "a", "op": "IN", "val": ["x"]}]}
+    )
+
+    screenshot_instance = MagicMock()
+    screenshot_instance.get_screenshot.return_value = b"PNGBYTES"
+
+    with (
+        patch(
+            f"{MODULE}.get_dashboard_filter_context", return_value=filter_context
+        ) as mock_ctx,
+        patch(f"{MODULE}.get_url_path", return_value="/explore/url") as mock_url,
+        patch(
+            f"{MODULE}.ChartScreenshot", return_value=screenshot_instance
+        ) as mock_screenshot,
+        patch.object(screenshot_module, "current_app", _patched_app()),
+    ):
+        user = MagicMock()
+        result = render_chart_image(
+            chart, dashboard_id=7, active_data_mask=active_mask, user=user
+        )
+
+    assert result == b"PNGBYTES"
+
+    # Live filter state is resolved for this chart on this dashboard.
+    mock_ctx.assert_called_once_with(
+        dashboard_id=7, chart_id=42, active_data_mask=active_mask
+    )
+
+    # The Explore URL carries the slice id and the live extra_form_data.
+    _, url_kwargs = mock_url.call_args
+    form_data = json.loads(url_kwargs["form_data"])
+    assert form_data["slice_id"] == 42
+    assert form_data["viz_type"] == "pie"
+    assert form_data["extra_form_data"] == filter_context.extra_form_data
+
+    # ChartScreenshot is built with the chart digest + slice window sizing.
+    args, kwargs = mock_screenshot.call_args
+    assert args[0] == "/explore/url"
+    assert args[1] == "abc123"
+    assert kwargs["window_size"] == (3000, 1200)
+    assert kwargs["thumb_size"] == (3000, 1200)
+    screenshot_instance.get_screenshot.assert_called_once_with(user=user)
+
+
+def test_render_chart_image_omits_extra_form_data_when_no_filters() -> None:
+    chart = _chart()
+    filter_context = DashboardFilterContext(extra_form_data={})
+    screenshot_instance = MagicMock()
+    screenshot_instance.get_screenshot.return_value = b"PNG"
+
+    with (
+        patch(f"{MODULE}.get_dashboard_filter_context", return_value=filter_context),
+        patch(f"{MODULE}.get_url_path", return_value="/u") as mock_url,
+        patch(f"{MODULE}.ChartScreenshot", return_value=screenshot_instance),
+        patch.object(screenshot_module, "current_app", _patched_app()),
+    ):
+        render_chart_image(chart, dashboard_id=7, active_data_mask={}, user=MagicMock())
+
+    _, url_kwargs = mock_url.call_args
+    form_data = json.loads(url_kwargs["form_data"])
+    assert "extra_form_data" not in form_data
+
+
+def test_render_chart_image_returns_none_when_screenshot_fails() -> None:
+    chart = _chart()
+    screenshot_instance = MagicMock()
+    screenshot_instance.get_screenshot.return_value = None
+
+    with (
+        patch(
+            f"{MODULE}.get_dashboard_filter_context",
+            return_value=DashboardFilterContext(),
+        ),
+        patch(f"{MODULE}.get_url_path", return_value="/u"),
+        patch(f"{MODULE}.ChartScreenshot", return_value=screenshot_instance),
+        patch.object(screenshot_module, "current_app", _patched_app()),
+    ):
+        result = render_chart_image(
+            chart, dashboard_id=7, active_data_mask={}, user=MagicMock()
+        )
+
+    assert result is None
+
+
+def test_render_chart_image_returns_none_on_exception() -> None:
+    chart = _chart()
+
+    with patch(
+        f"{MODULE}.get_dashboard_filter_context", side_effect=ValueError("boom")
+    ):
+        result = render_chart_image(
+            chart, dashboard_id=7, active_data_mask={}, user=MagicMock()
+        )
+
+    assert result is None
+
+
+def test_render_chart_image_propagates_soft_time_limit() -> None:
+    # A soft timeout must abort the export, not be swallowed into a ``None``
+    # result (which the caller would mis-report as a per-chart render failure).
+    chart = _chart()
+
+    with (
+        patch(
+            f"{MODULE}.get_dashboard_filter_context",
+            side_effect=SoftTimeLimitExceeded(),
+        ),
+        pytest.raises(SoftTimeLimitExceeded),
+    ):
+        render_chart_image(chart, dashboard_id=7, active_data_mask={}, user=MagicMock())
diff --git a/tests/unit_tests/tasks/test_export_dashboard_excel.py b/tests/unit_tests/tasks/test_export_dashboard_excel.py
new file mode 100644
index 00000000000..013a8ebf8cf
--- /dev/null
+++ b/tests/unit_tests/tasks/test_export_dashboard_excel.py
@@ -0,0 +1,401 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+import glob
+import os
+import tempfile
+from collections.abc import Iterator
+from contextlib import ExitStack
+from typing import Any
+from unittest import mock
+
+import pytest
+from celery.exceptions import SoftTimeLimitExceeded
+
+from superset.utils import json
+
+MODULE = "superset.tasks.export_dashboard_excel"
+
+
+# A minimal valid 1x1 transparent PNG for image-mode tests.
+_PNG_1x1 = (
+    b"\x89PNG\r\n\x1a\n\x00\x00\x00\rIHDR\x00\x00\x00\x01\x00\x00\x00\x01\x08\x06"
+    b"\x00\x00\x00\x1f\x15\xc4\x89\x00\x00\x00\rIDATx\x9cc\x00\x01\x00\x00\x05\x00"
+    b"\x01\r\n-\xb4\x00\x00\x00\x00IEND\xaeB`\x82"
+)
+
+
+def _chart(
+    chart_id: int,
+    name: str,
+    has_context: bool = True,
+    viz_type: str = "line",
+) -> mock.MagicMock:
+    chart = mock.MagicMock()
+    chart.id = chart_id
+    chart.slice_name = name
+    chart.viz_type = viz_type
+    chart.query_context = json.dumps({"queries": [{}]}) if has_context else None
+    return chart
+
+
+def _media(path: str) -> list[str]:
+    """Embedded media entries of an xlsx (which is a zip archive)."""
+    import zipfile
+
+    with zipfile.ZipFile(path) as archive:
+        return [n for n in archive.namelist() if n.startswith("xl/media/")]
+
+
+@pytest.fixture
+def mocks() -> Iterator[dict[str, Any]]:
+    """Patch every external dependency of the task; keep the real xlsx writer."""
+    with ExitStack() as stack:
+        # Use explicit MagicMock instances: patch() auto-creates async-flavored
+        # mocks for these targets (their real objects expose async members), which
+        # would make calls like security_manager.get_user_by_id() return coroutines.
+        patched = {
+            name: stack.enter_context(
+                mock.patch(f"{MODULE}.{name}", new=mock.MagicMock())
+            )
+            for name in (
+                "security_manager",
+                "db",
+                "get_charts_in_layout_order",
+                "get_dashboard_filter_context",
+                "ChartDataQueryContextSchema",
+                "ChartDataCommand",
+                "render_chart_image",
+                "s3",
+                "email",
+                "ReleaseDistributedLock",
+            )
+        }
+        user = mock.MagicMock()
+        user.email = "user@example.com"
+        patched["security_manager"].get_user_by_id.return_value = user
+
+        dashboard = mock.MagicMock()
+        dashboard.id = 1
+        dashboard.dashboard_title = "Sales"
+        patched[
+            "db"
+        ].session.query.return_value.filter_by.return_value.one_or_none.return_value = (  # noqa: E501
+            dashboard
+        )
+
+        patched["get_dashboard_filter_context"].return_value.extra_form_data = {}
+        patched["s3"].generate_presigned_url.return_value = "https://signed/file.xlsx"
+
+        patched["user"] = user
+        patched["dashboard"] = dashboard
+        yield patched
+
+
+def _run(
+    job_id: str = "job-1",
+    mode: str = "data",
+) -> None:
+    from superset.tasks.export_dashboard_excel import export_dashboard_excel
+
+    export_dashboard_excel(
+        dashboard_id=1,
+        user_id=2,
+        active_data_mask={},
+        job_id=job_id,
+        mode=mode,
+    )
+
+
+def _no_temp_files_left(job_id: str) -> bool:
+    pattern = os.path.join(tempfile.gettempdir(), f"dash-export-{job_id}-*")
+    return glob.glob(pattern) == []
+
+
+def _read_sheets(path: str) -> dict[str, list[list[object]]]:
+    openpyxl = pytest.importorskip("openpyxl")
+    workbook = openpyxl.load_workbook(path, read_only=True)
+    sheets = {
+        ws.title: [list(r) for r in ws.iter_rows(values_only=True)]
+        for ws in workbook.worksheets
+    }
+    workbook.close()
+    return sheets
+
+
+def test_happy_path_uploads_and_emails(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "First"),
+        _chart(20, "Second"),
+    ]
+    mocks["ChartDataCommand"].return_value.run.side_effect = [
+        {"queries": [{"colnames": ["a", "b"], "data": [{"a": 1, "b": 2}]}]},
+        {"queries": [{"colnames": ["c"], "data": [{"c": "x"}]}]},
+    ]
+
+    # Capture the workbook before the task deletes it.
+    uploaded: dict[str, Any] = {}
+
+    def _capture(path: str, bucket: str, key: str) -> None:
+        uploaded["sheets"] = _read_sheets(path)
+
+    mocks["s3"].upload_file_to_s3.side_effect = _capture
+
+    _run()
+
+    mocks["s3"].upload_file_to_s3.assert_called_once()
+    assert list(uploaded["sheets"].keys()) == ["10 - First", "20 - Second"]
+    mocks["email"].send_export_email.assert_called_once()
+    mocks["email"].build_success_email.assert_called_once()
+    assert _no_temp_files_left("job-1")
+
+
+def test_chart_without_query_context_is_skipped(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Good"),
+        _chart(20, "NoContext", has_context=False),
+    ]
+    mocks["ChartDataCommand"].return_value.run.return_value = {
+        "queries": [{"colnames": ["a"], "data": [{"a": 1}]}]
+    }
+
+    _run()
+
+    _, kwargs = mocks["email"].build_success_email.call_args
+    assert kwargs["errored"] == {
+        mocks["email"].ERROR_NO_QUERY_CONTEXT: ["20 - NoContext"]
+    }
+
+
+def test_chart_query_error_grouped_as_general_export_continues(
+    mocks: dict[str, Any],
+) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Boom"),
+        _chart(20, "Ok"),
+    ]
+    mocks["ChartDataCommand"].return_value.run.side_effect = [
+        RuntimeError("query failed"),
+        {"queries": [{"colnames": ["a"], "data": [{"a": 1}]}]},
+    ]
+
+    _run()
+
+    mocks["s3"].upload_file_to_s3.assert_called_once()
+    _, kwargs = mocks["email"].build_success_email.call_args
+    assert kwargs["errored"] == {mocks["email"].ERROR_GENERAL: ["10 - Boom"]}
+
+
+def test_chart_timeout_aborts_export_and_sends_failure_email(
+    mocks: dict[str, Any],
+) -> None:
+    # A soft timeout raised while a chart runs must abort the whole export
+    # (propagate to the outer handler) rather than being recorded per-chart and
+    # letting the task run on until the hard limit kills the worker — which would
+    # skip cleanup, leak temp files, and never send a failure email.
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Ok"),
+        _chart(20, "Slow"),
+    ]
+    mocks["ChartDataCommand"].return_value.run.side_effect = [
+        {"queries": [{"colnames": ["a"], "data": [{"a": 1}]}]},
+        SoftTimeLimitExceeded(),
+    ]
+
+    with pytest.raises(SoftTimeLimitExceeded):
+        _run("job-timeout")
+
+    mocks["s3"].upload_file_to_s3.assert_not_called()
+    mocks["email"].build_success_email.assert_not_called()
+    mocks["email"].build_failure_email.assert_called_once()
+    assert _no_temp_files_left("job-timeout")
+
+
+def test_image_render_timeout_aborts_export(mocks: dict[str, Any]) -> None:
+    # In image mode a soft timeout during a chart render must also propagate and
+    # abort the export rather than being swallowed as a per-chart render failure.
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Line", viz_type="line"),
+    ]
+    mocks["render_chart_image"].side_effect = SoftTimeLimitExceeded()
+
+    with pytest.raises(SoftTimeLimitExceeded):
+        _run("job-img-timeout", mode="images")
+
+    mocks["email"].build_success_email.assert_not_called()
+    mocks["email"].build_failure_email.assert_called_once()
+    assert _no_temp_files_left("job-img-timeout")
+
+
+def test_all_charts_skipped_writes_summary(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "NoContext", has_context=False),
+    ]
+    uploaded: dict[str, Any] = {}
+
+    def _capture(path: str, bucket: str, key: str) -> None:
+        uploaded["sheets"] = _read_sheets(path)
+
+    mocks["s3"].upload_file_to_s3.side_effect = _capture
+
+    _run()
+
+    assert "Export Summary" in uploaded["sheets"]
+    mocks["email"].build_success_email.assert_called_once()
+
+
+def test_upload_failure_sends_failure_email_and_cleans_up(
+    mocks: dict[str, Any],
+) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [_chart(10, "Good")]
+    mocks["ChartDataCommand"].return_value.run.return_value = {
+        "queries": [{"colnames": ["a"], "data": [{"a": 1}]}]
+    }
+    mocks["s3"].upload_file_to_s3.side_effect = RuntimeError("s3 down")
+
+    with pytest.raises(RuntimeError):
+        _run("job-fail")
+
+    mocks["email"].build_failure_email.assert_called_once()
+    mocks["email"].send_export_email.assert_called_once()
+    assert _no_temp_files_left("job-fail")
+
+
+def test_soft_time_limit_sends_failure_email(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].side_effect = SoftTimeLimitExceeded()
+
+    with pytest.raises(SoftTimeLimitExceeded):
+        _run("job-timeout")
+
+    mocks["email"].build_failure_email.assert_called_once()
+    assert _no_temp_files_left("job-timeout")
+
+
+# --- image mode ---
+
+
+def test_images_mode_embeds_non_table_and_keeps_tables_tabular(
+    mocks: dict[str, Any],
+) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Line", viz_type="line"),
+        _chart(20, "Tbl", viz_type="table"),
+    ]
+    mocks["render_chart_image"].return_value = _PNG_1x1
+    mocks["ChartDataCommand"].return_value.run.return_value = {
+        "queries": [{"colnames": ["a"], "data": [{"a": 1}]}]
+    }
+
+    uploaded: dict[str, Any] = {}
+
+    def _capture(path: str, bucket: str, key: str) -> None:
+        uploaded["sheets"] = _read_sheets(path)
+        uploaded["media"] = _media(path)
+
+    mocks["s3"].upload_file_to_s3.side_effect = _capture
+
+    _run(mode="images")
+
+    # The non-table chart is rendered as an image (with the requesting user)...
+    mocks["render_chart_image"].assert_called_once()
+    render_args = mocks["render_chart_image"].call_args[0]
+    assert render_args[0].id == 10
+    assert render_args[3] is mocks["user"]
+    # ...and the table chart still goes through the data path.
+    mocks["ChartDataCommand"].return_value.run.assert_called_once()
+
+    assert set(uploaded["sheets"].keys()) == {"10 - Line", "20 - Tbl"}
+    # Exactly one embedded image (the non-table chart).
+    assert len(uploaded["media"]) == 1
+
+
+def test_images_mode_renders_chart_without_query_context(
+    mocks: dict[str, Any],
+) -> None:
+    # An image chart with no saved query context can still render.
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Line", viz_type="line", has_context=False),
+    ]
+    mocks["render_chart_image"].return_value = _PNG_1x1
+
+    uploaded: dict[str, Any] = {}
+
+    def _capture(path: str, bucket: str, key: str) -> None:
+        uploaded["media"] = _media(path)
+
+    mocks["s3"].upload_file_to_s3.side_effect = _capture
+
+    _run(mode="images")
+
+    mocks["render_chart_image"].assert_called_once()
+    assert len(uploaded["media"]) == 1
+
+
+def test_images_mode_none_render_is_skipped(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [
+        _chart(10, "Line", viz_type="line"),
+    ]
+    mocks["render_chart_image"].return_value = None
+
+    uploaded: dict[str, Any] = {}
+
+    def _capture(path: str, bucket: str, key: str) -> None:
+        uploaded["sheets"] = _read_sheets(path)
+
+    mocks["s3"].upload_file_to_s3.side_effect = _capture
+
+    _run(mode="images")
+
+    # A chart that cannot render is grouped under the general-error reason.
+    _, kwargs = mocks["email"].build_success_email.call_args
+    assert kwargs["errored"] == {mocks["email"].ERROR_GENERAL: ["10 - Line"]}
+    # Nothing rendered → the summary sheet stands in for an empty workbook.
+    assert "Export Summary" in uploaded["sheets"]
+
+
+def test_inflight_lock_released_on_success(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [_chart(10, "Good")]
+    mocks["ChartDataCommand"].return_value.run.return_value = {
+        "queries": [{"colnames": ["a"], "data": [{"a": 1}]}]
+    }
+
+    _run()
+
+    # The distributed lock is released for this user+dashboard when the task
+    # settles (namespace + params match what the API acquired).
+    mocks["ReleaseDistributedLock"].assert_called_once_with(
+        "excel_export", {"user_id": 2, "dashboard_id": 1}
+    )
+    mocks["ReleaseDistributedLock"].return_value.run.assert_called_once_with()
+
+
+def test_inflight_lock_released_on_failure(mocks: dict[str, Any]) -> None:
+    mocks["get_charts_in_layout_order"].return_value = [_chart(10, "Good")]
+    mocks["ChartDataCommand"].return_value.run.return_value = {
+        "queries": [{"colnames": ["a"], "data": [{"a": 1}]}]
+    }
+    mocks["s3"].upload_file_to_s3.side_effect = RuntimeError("s3 down")
+
+    with pytest.raises(RuntimeError):
+        _run("job-fail")
+
+    # The lock is freed in ``finally`` even when the export fails.
+    mocks["ReleaseDistributedLock"].assert_called_once_with(
+        "excel_export", {"user_id": 2, "dashboard_id": 1}
+    )
+    mocks["ReleaseDistributedLock"].return_value.run.assert_called_once_with()
diff --git a/tests/unit_tests/utils/excel_streaming_tests.py b/tests/unit_tests/utils/excel_streaming_tests.py
new file mode 100644
index 00000000000..822db329d8b
--- /dev/null
+++ b/tests/unit_tests/utils/excel_streaming_tests.py
@@ -0,0 +1,267 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+from datetime import date, datetime
+from decimal import Decimal
+from pathlib import Path
+
+import pytest
+
+from superset.utils import excel_streaming
+from superset.utils.excel_streaming import (
+    _sanitize_cell,
+    sanitize_sheet_name,
+    StreamingXlsxWriter,
+)
+
+# --- sanitize_sheet_name ---
+
+
+def test_sheet_name_replaces_forbidden_chars() -> None:
+    assert sanitize_sheet_name("a/b:c*d?e[f]g\\h", set()) == "a_b_c_d_e_f_g_h"
+
+
+def test_sheet_name_truncated_to_31() -> None:
+    assert sanitize_sheet_name("x" * 40, set()) == "x" * 31
+
+
+def test_sheet_name_dedupes_case_insensitively() -> None:
+    used: set[str] = set()
+    assert sanitize_sheet_name("Sales", used) == "Sales"
+    assert sanitize_sheet_name("sales", used) == "sales~2"
+    assert sanitize_sheet_name("SALES", used) == "SALES~3"
+
+
+def test_sheet_name_dedupe_marker_respects_length_cap() -> None:
+    used: set[str] = set()
+    long_name = "y" * 31
+    assert sanitize_sheet_name(long_name, used) == long_name
+    assert sanitize_sheet_name(long_name, used) == "y" * 29 + "~2"
+
+
+def test_sheet_name_blank_falls_back() -> None:
+    assert sanitize_sheet_name("   ", set()) == "Sheet"
+
+
+def test_sheet_name_reserved_history_is_escaped() -> None:
+    assert sanitize_sheet_name("History", set()) == "History_"
+
+
+def test_sheet_name_strips_surrounding_apostrophes() -> None:
+    assert sanitize_sheet_name("'quoted'", set()) == "quoted"
+
+
+# --- _sanitize_cell ---
+
+
+@pytest.mark.parametrize(
+    "value,expected",
+    [
+        (None, ""),
+        ("=SUM(A1)", "'=SUM(A1)"),
+        ("+1", "'+1"),
+        ("-1", "'-1"),
+        ("@handle", "'@handle"),
+        ("normal", "normal"),
+        (True, True),
+        (5, 5),
+        (1.5, 1.5),
+        (Decimal("2.5"), 2.5),
+        (datetime(2020, 1, 2, 3, 4, 5), "2020-01-02T03:04:05"),
+        (date(2020, 1, 2), "2020-01-02"),
+    ],
+)
+def test_sanitize_cell(value: object, expected: object) -> None:
+    assert _sanitize_cell(value) == expected
+
+
+def test_sanitize_cell_large_int_becomes_string() -> None:
+    assert _sanitize_cell(10**16) == str(10**16)
+
+
+def test_sanitize_cell_non_finite_floats_blanked() -> None:
+    assert _sanitize_cell(float("nan")) == ""
+    assert _sanitize_cell(float("inf")) == ""
+
+
+def test_sanitize_cell_non_finite_decimals_blanked() -> None:
+    # float(Decimal("NaN")) is nan and float(Decimal("Infinity")) is inf, both
+    # of which xlsxwriter rejects; they must be blanked rather than crash.
+    assert _sanitize_cell(Decimal("NaN")) == ""
+    assert _sanitize_cell(Decimal("Infinity")) == ""
+    assert _sanitize_cell(Decimal("-Infinity")) == ""
+
+
+def test_sanitize_cell_out_of_range_decimal_is_blanked() -> None:
+    # A Decimal too large for a float becomes inf (or, in edge cases, raises
+    # OverflowError); either way it must be neutralized rather than crash
+    # xlsxwriter or emit a bogus value.
+    assert _sanitize_cell(Decimal("1E10000")) == ""
+
+
+@pytest.mark.parametrize(
+    "value,expected",
+    [
+        (" =SUM(A1)", "' =SUM(A1)"),
+        ("\t=SUM(A1)", "'\t=SUM(A1)"),
+        ("  +1", "'  +1"),
+        ("\t@handle", "'\t@handle"),
+    ],
+)
+def test_sanitize_cell_quotes_formula_behind_whitespace(
+    value: str, expected: str
+) -> None:
+    # Spreadsheet apps evaluate formulas even when preceded by spaces/tabs, so
+    # the formula guard must look past leading whitespace.
+    assert _sanitize_cell(value) == expected
+
+
+# --- StreamingXlsxWriter (round-trip via openpyxl) ---
+
+
+def _read_workbook(path: str) -> dict[str, list[list[object]]]:
+    openpyxl = pytest.importorskip("openpyxl")
+    workbook = openpyxl.load_workbook(path, read_only=True)
+    sheets = {
+        ws.title: [list(row) for row in ws.iter_rows(values_only=True)]
+        for ws in workbook.worksheets
+    }
+    workbook.close()
+    return sheets
+
+
+def test_writer_writes_one_sheet_per_chart(tmp_path: Path) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    assert writer.add_sheet("10 - First", ["a", "b"], [[1, 2], [3, 4]]) == 2
+    assert writer.add_sheet("20 - Second", ["c"], [["x"]]) == 1
+    writer.close()
+
+    sheets = _read_workbook(path)
+    assert list(sheets.keys()) == ["10 - First", "20 - Second"]
+    assert sheets["10 - First"] == [["a", "b"], [1, 2], [3, 4]]
+    assert sheets["20 - Second"] == [["c"], ["x"]]
+
+
+def test_writer_quotes_formula_cells(tmp_path: Path) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    writer.add_sheet("data", ["col"], [["=cmd()"]])
+    writer.close()
+
+    sheets = _read_workbook(path)
+    assert sheets["data"][1][0] == "'=cmd()"
+
+
+def test_writer_caps_rows_per_sheet(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    monkeypatch.setattr(excel_streaming, "MAX_DATA_ROWS_PER_SHEET", 3)
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    written = writer.add_sheet("data", ["col"], [[i] for i in range(5)])
+    writer.close()
+
+    # One row is reserved for the truncation notice, so only 2 data rows fit.
+    assert written == 2
+    sheets = _read_workbook(path)
+    # header + 2 data rows + 1 truncation notice
+    assert len(sheets["data"]) == 4
+    assert sheets["data"][-1][0] == "[Truncated: only first 2 rows exported]"
+
+
+def test_writer_no_truncation_notice_when_data_fits(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    monkeypatch.setattr(excel_streaming, "MAX_DATA_ROWS_PER_SHEET", 3)
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    # Exactly fills the reserved capacity (MAX - 1) with no leftover rows.
+    written = writer.add_sheet("data", ["col"], [[i] for i in range(2)])
+    writer.close()
+
+    assert written == 2
+    sheets = _read_workbook(path)
+    # header + 2 data rows, no notice
+    assert sheets["data"] == [["col"], [0], [1]]
+
+
+def test_writer_empty_workbook_is_valid(tmp_path: Path) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    writer.close()
+
+    sheets = _read_workbook(path)
+    assert list(sheets.keys()) == ["Export"]
+
+
+def test_writer_summary_sheet(tmp_path: Path) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    writer.add_summary_sheet("Export Summary", ["Skipped charts:", "10 - Broken"])
+    writer.close()
+
+    sheets = _read_workbook(path)
+    assert sheets["Export Summary"] == [["Skipped charts:"], ["10 - Broken"]]
+
+
+# --- add_image_sheet ---
+
+# A minimal valid 1x1 transparent PNG.
+_PNG_1x1 = (
+    b"\x89PNG\r\n\x1a\n\x00\x00\x00\rIHDR\x00\x00\x00\x01\x00\x00\x00\x01\x08\x06"
+    b"\x00\x00\x00\x1f\x15\xc4\x89\x00\x00\x00\rIDATx\x9cc\x00\x01\x00\x00\x05\x00"
+    b"\x01\r\n-\xb4\x00\x00\x00\x00IEND\xaeB`\x82"
+)
+
+
+def _read_media(path: str) -> list[str]:
+    """Return the embedded media entries of an xlsx (which is a zip archive)."""
+    import zipfile
+
+    with zipfile.ZipFile(path) as archive:
+        return [n for n in archive.namelist() if n.startswith("xl/media/")]
+
+
+def test_add_image_sheet_embeds_image_and_counts(tmp_path: Path) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    writer.add_image_sheet("10 - Chart", _PNG_1x1)
+    assert writer.sheet_count == 1
+    writer.close()
+
+    # The sheet exists (with a sanitized/unique name) and the PNG is embedded.
+    sheets = _read_workbook(path)
+    assert list(sheets.keys()) == ["10 - Chart"]
+    assert _read_media(path) == ["xl/media/image1.png"]
+
+
+def test_add_image_sheet_dedupes_and_composes_with_data_sheets(
+    tmp_path: Path,
+) -> None:
+    path = str(tmp_path / "out.xlsx")
+    writer = StreamingXlsxWriter(path)
+    writer.add_sheet("10 - Chart", ["a"], [[1]])
+    writer.add_image_sheet("10 - Chart", _PNG_1x1)
+    writer.close()
+
+    assert writer.sheet_count == 2
+    sheets = _read_workbook(path)
+    # The image sheet name is de-duplicated against the existing data sheet.
+    assert list(sheets.keys()) == ["10 - Chart", "10 - Chart~2"]
+    assert _read_media(path) == ["xl/media/image1.png"]
diff --git a/tests/unit_tests/utils/s3_tests.py b/tests/unit_tests/utils/s3_tests.py
new file mode 100644
index 00000000000..7b9926f3a63
--- /dev/null
+++ b/tests/unit_tests/utils/s3_tests.py
@@ -0,0 +1,95 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+from __future__ import annotations
+
+import sys
+from unittest.mock import MagicMock, patch
+
+import pytest
+
+from superset.utils import s3
+
+
+@patch("boto3.client")
+@patch("superset.utils.s3.current_app")
+def test_upload_file_to_s3(mock_app: MagicMock, mock_client_fn: MagicMock) -> None:
+    mock_app.config = {"EXCEL_EXPORT_S3_CLIENT_KWARGS": {}}
+    client = mock_client_fn.return_value
+
+    s3.upload_file_to_s3("exports/out.xlsx", "my-bucket", "exports/1/abc.xlsx")
+
+    mock_client_fn.assert_called_once_with("s3")
+    client.upload_file.assert_called_once_with(
+        "exports/out.xlsx", "my-bucket", "exports/1/abc.xlsx"
+    )
+
+
+@patch("boto3.client")
+@patch("superset.utils.s3.current_app")
+def test_client_kwargs_passthrough(
+    mock_app: MagicMock, mock_client_fn: MagicMock
+) -> None:
+    mock_app.config = {
+        "EXCEL_EXPORT_S3_CLIENT_KWARGS": {
+            "endpoint_url": "http://minio:9000",
+            "region_name": "us-east-1",
+        }
+    }
+
+    s3.upload_file_to_s3("exports/out.xlsx", "my-bucket", "k")
+
+    mock_client_fn.assert_called_once_with(
+        "s3", endpoint_url="http://minio:9000", region_name="us-east-1"
+    )
+
+
+def test_importing_module_does_not_require_boto3() -> None:
+    # Regression: importing this module (which app startup does via the dashboard
+    # API) must not require boto3, since it is only an optional install.
+    import importlib
+
+    with patch.dict(sys.modules, {"boto3": None}):
+        importlib.reload(s3)
+    # Reload again with boto3 available so later tests see the normal module.
+    importlib.reload(s3)
+
+
+def test_get_s3_client_missing_boto3_raises_actionable_error() -> None:
+    # Simulate a production install without boto3: the lazy import fails and we
+    # surface an install hint instead of a bare ModuleNotFoundError.
+    with (
+        patch.dict(sys.modules, {"boto3": None}),
+        pytest.raises(ImportError, match="excel-export"),
+    ):
+        s3._get_s3_client()
+
+
+@patch("boto3.client")
+@patch("superset.utils.s3.current_app")
+def test_generate_presigned_url(mock_app: MagicMock, mock_client_fn: MagicMock) -> None:
+    mock_app.config = {"EXCEL_EXPORT_S3_CLIENT_KWARGS": {}}
+    client = mock_client_fn.return_value
+    client.generate_presigned_url.return_value = "https://signed.example/abc"
+
+    url = s3.generate_presigned_url("my-bucket", "exports/1/abc.xlsx", 86400)
+
+    assert url == "https://signed.example/abc"
+    client.generate_presigned_url.assert_called_once_with(
+        "get_object",
+        Params={"Bucket": "my-bucket", "Key": "exports/1/abc.xlsx"},
+        ExpiresIn=86400,
+    )