Mockup for approval. Compliee owns the engine per the boundary ruling; boommedia.us is client zero.
Sample data is illustrative — but every status shown is a real condition found on your estate today.
compliee.online/monitor
Estate overview
23 monitored endpoints · 14 apps · 9 client sites · updated 2 min ago
Estate health
22/23
1 endpoint down — rankee.online (apex)
▲ Action needed
Uptime
99.94%
7-day, all endpoints
Attacks blocked
1,204
+18% vs last week
Certs expiring
2
within 30 days
Failed deploys
3
production, 7 days
Median response
218ms
across all sites
Unrun migrations
4
across 2 apps
Attacks blocked
Requests stopped at the edge or by middleware, per day
Blocked requests
What was blocked
By attack pattern, last 7 days
Response time
Median time to first byte, 7 days
Boom MediaClient sites (median)
Every endpoint
Sorted by what needs attention
Endpoint
Status
Uptime 7d
Response
Certificate
Blocked
rankee.online(apex)
✕ Down
0%
—
✕ Untrusted
—
untouchables.…
✓ Up
100%
241ms
▲ 11 days
96
guarapos.boomonlineordering.com
✓ Up
99.98%
198ms
✓ 74 days
312
www.boommedia.us
✓ Up
100%
212ms
✓ 68 days
204
db.boommedia.us(shared)
✓ Up
99.91%
88ms
✓ 51 days
—
Why it looks like this
Status is never colour alone. Every pill carries an icon and a word — ✓ Up, ▲ 11 days,
✕ Down. A red-green dashboard is unreadable to roughly one man in twelve, and this one goes in front of
clients.
The colours were computed, not chosen. The two series colours were run through a
contrast and colour-vision validator against this exact surface: worst-pair ΔE 19.4 under deuteranopia,
26.5 normal vision, both above 3:1 contrast. The status trio separates at ΔE 11.3 CVD / 27.6 normal.
One axis, one hue per series, no dual axes. Response time and attack volume are
different scales, so they are different charts rather than two y-axes on one plot — the single most
common way a dashboard invents a correlation that is not in the data.
Numbers appear where they earn their place. Peak and trough are labelled on the bars,
series are labelled at the line ends. A value on every point is noise, and the table underneath is
where exact figures live.
Every status shown is real
The sample data is illustrative, but the conditions are not invented — they are what is
actually true on your estate right now, which is the point of building this:
rankee.online apex down — 503 behind a Traefik default certificate, verified today
Failed production deploys — several ERROR-state deploys sit in the Vercel history, unnoticed
Unrun migrations — four surfaced in Rankee alone in one session
Certificate expiry — nothing watches this today
What it takes to make this real
The charts are the easy part; the feeds are the work. Uptime and response time come
from UptimeRobot — whose API is paid-tier only ($9/mo), so this dashboard is the one
thing in the plan that is not free. Attack data needs Phases A–D of the Compliee spec (CrowdSec +
@boom/shield + ingest). Deploy status is the Vercel API, already available. Migration
status is the Rankee pattern ported to each app.
Build order stands: feeds first, dashboard last. A dashboard built before the data
exists is a mockup you rewrite. This page is that mockup — approve the shape, then we wire it as each
feed lands.