Live dashboard

A live dev-server dashboard. Active SSE connections, per-intent counters, recent telemetry events, registry sizes — all in one auto-refreshing screen.

The dashboard is itself a Wun screen (defscreen :wun.dashboard/main), so it dogfoods the framework: same patch pipeline, same wire envelope, same SSE channel as user screens. There is no privileged path.

Mount it

The framework does not auto-mount. From your app’s server init:

(ns myapp.server.init
  (:require [wun.server.dashboard :as dashboard]))

(defn start! []
  (dashboard/install!)
  ;; ... start your Pedestal service ...
  )

Then visit http://localhost:8080/_wun/dashboard.

install! is idempotent — calling it more than once on the same JVM is a no-op.

Production posture

Dashboard mounting is refused when WUN_PROFILE=prod. To opt-in on a production host (for example, behind a VPN-only admin route), set both:

WUN_PROFILE=prod WUN_DASHBOARD_FORCE=1

The dashboard exposes per-connection state, recent telemetry events, and the full intent vocabulary — useful for debugging, not something to expose to the open internet by default.

What it shows

Section	Source
Active connections	@wun.server.state/connections — conn-id, screen, caps
Intent metrics	rolling counters per :wun/intent.applied / .rejected
Recent telemetry	last 200 events, all event keys
Registry sizes	defcomponent / defscreen / defintent counts
Uptime	wall time since install! was called

The render fn is pure and reads everything from per-connection state under :wun.dashboard/snapshot. A 1 Hz server-side ticker walks active dashboard connections, swaps a fresh snapshot into each, and triggers a normal broadcast — so the dashboard updates through the same patch pipeline as any other screen.

Limits

This is a live snapshot tool, not a metrics backend.

• No persistence: events and counters reset when the JVM restarts.
• No sliding windows or time-series buckets — counters are cumulative since install!.
• Read-only: no “kick this connection” or live state editing.
• No charts. Tables and badges only.

For long-horizon metrics, register your own wun.server.telemetry/register-sink! that forwards to Prometheus / OpenTelemetry / your aggregator of choice. The dashboard’s tap is one such sink — read its source for a template.