The administrator landing page — system overview, security metrics, service health, detection charts and the newest security and audit events.

The dashboard is the landing page for administrative roles. It answers three questions in one screen: how much the appliance is being used, what its detection pipeline is catching, and whether its services are healthy.

What do I need?

Licence: Any
Role: admin_userviewer

Prerequisites

You are signed in as an admin_user, the superadmin, or a viewer. A normal_user is sent to the chat instead, and an auditor to Compliance.

The subtitle under the title states the scope you are looking at — Security overview, then the time range, then all tenants.

Screenshot pendingDashboard with overview tiles, security metrics, health, detection charts and the security and audit table/assets/screenshots/dashboard@0.4.png

Choose the time range

Open the range menu

Select Time range in the page header.

Four ranges are offered: last 24 hours, last 7 days, last 30 days and all time.

Pick one

Select a range.

The ranged cards re-read against it and the subtitle updates. Three figures are not windowed and do not move: Total Tenants, Compliance, and the whole Recent activity card. The choice is remembered in this browser and is last 30 days until you change it.

Note

The Compliance figure on Security Metrics is cumulative and ignores the range — it counts violations prevented over the life of the appliance.

Read the cards

The page ships with seven cards, in this order.

Card	What it answers
System Overview	Total Tenants across all time, then Active Sessions and API Requests for the range, with a link to Tenants
Security Metrics	Detections, Compliance violations prevented, and Threats Rejected — outright blocks, highlighted when above zero
System Health	A Healthy or Warning summary of the error rate, then a row per service: API gateway, ML detector (NER), Detections API and Event logging, each Healthy, Degraded or Unreachable
Detections by action	A donut splitting detections into Flagged, Masked and Rejected, with the prompt/response split underneath
Top entity types	Detections by entity type, split into PII types and Security, with the share of detections the top entries account for
Security & Audit	The newest policy, access and configuration events, with Category, Event, Detail, Actor and When
Recent activity	The newest detections, with User, Status, Entities, Source and When

Every card refreshes itself once a minute.

Follow a card to its page

Select the link at the foot of a card, or select a row.

Go to Tenants, Go to All Detections and Go to Event Logs open the corresponding page. Selecting a Security & Audit row opens Event logs already filtered to that category.

Rearrange the dashboard

Enter customise mode

Select Customise.

A banner explains the controls: Drag a card by its grip, or focus the grip and use the arrow keys. Use the eye to hide or show a card. Saved in this browser only.

Reorder and hide cards

Drag a card by its grip, or focus the grip and use the arrow keys. Each card's edit bar carries its own move and hide controls.

Outside customise mode the same three actions are on a card's own menu — Move up, Move down, Hide. That menu is not rendered while customise mode is on.

Hidden cards are counted in a Hidden cards dropdown with a Show all action.

Finish

Select Done, or press Escape.

The layout is stored in this browser only. It does not follow you to another device and it is not shared with other administrators.

To undo everything, select Reset to default and confirm Reset this dashboard? — Every card comes back in the order it ships in, and the arrangement you made is discarded. This cannot be undone.

Print or export the view

Select Export report. This opens your browser's print dialog on the current dashboard, so the output is whatever is on screen — the range you chose and the cards you left visible. It is not a generated CSV or PDF; for those, use the exports on Compliance reports and Token cost and savings.

Fields reference

Control	What it does
Time range	Switches every ranged card between last 24 hours, last 7 days, last 30 days and all time. Remembered per browser; default last 30 days
Export report	Opens the browser print dialog on the current view
Start chat	Opens the chat playground
Customise / Done	Enters and leaves layout editing
Reset to default	Restores the shipped card order and unhides every card
View all	Opens the full list behind a card

Tab	Opens
Overview	This page
All Detections	All detections
Compliance	Compliance
Event Logs	Event logs

Verify

The subtitle names the range you selected, and the figures change when you switch range.
System Health reports Healthy for all four services. Anything else names the service that is degraded or unreachable.
Send a message through the chat that your filters catch, wait for the next refresh, and confirm it appears in Recent activity — a masked message as Masked, a blocked one as Request rejected.
Security & Audit lists the events behind your recent administrative changes; a user edit shows as a RESOURCE_UPDATED row, a login as AUTH_SUCCESS.

If it fails

The page redirects to the chat as soon as you open it — your role is normal_user, which has no dashboard. This is the registry's decision, not an error.
No recent activity — no detection has been recorded at all. This card is not windowed by the time range, so widening the range will not populate it.
No security or audit events in the {range} — the same, for the audit feed.
No detection data on Top entity types — no detections in the range to chart.
Warning on System Health with an error percentage — the event log holds errors in this window. Open Event logs and filter to the error category.
A service row reads Unreachable — that container is down or unhealthy. Check the service and its logs before trusting any figure that depends on it.
Cards show stale numbers — the page refreshes once a minute; reload if you need the answer now.

Next steps

All detections — the full feed behind the detection cards.
Event logs — the audit trail behind Security & Audit.
Tenant analytics — the same usage figures broken down per account.

Read the dashboard