Prometheus Intelligence · monitoring
Raw metrics in. Fleet intelligence out.
Prometheus already scrapes everything; this turns the firehose into answers.
Pay once in the store, paste your licence key in Settings, and the Metrics tab appears in QuoxCORE’s NOC panel. No subscription, ever.
In plain words
What it is, where it lives, when to reach for it
- What is it
- A plugin for the QuoxCORE dashboard that connects to the Prometheus server you already run and reads its metrics.
- Where do I use it
- On the Metrics tab in the dashboard's NOC panel, querying your own Prometheus over its API.
- When would I use it
- When you want to query metrics or check an alert without leaving the dashboard your agents work in.
- How do I use it
- Buy it once in the store for $19, then an org admin installs it from the dashboard's Plugins page.
QuoxCORE is the free, self-hosted platform underneath this. What is QuoxCORE
The Metrics tab, running on this page.
The plane above is the plugin’s idea of your fleet: every node a host, load as heat, anomalies flaring red with no alert rules written.
scraping- ok api-latency-p99 back under threshold
- warn disk-pressure · node-11 · 82% used
What it does
Answers, not just graphs.
Grafana shows you the dashboard you built. Prometheus Intelligence looks at the metrics you did not think to chart: it maps every host onto one heatmap, flags anomalies without a single alert rule, and projects capacity so a full disk is a calendar entry, not a 3am page.
For the times you want the raw material, the PromQL explorer runs instant and range queries with presets, and the alert timeline keeps severity history in one place.
Fleet health heatmap
Every monitored host as a cell, load as heat. The whole estate at a glance.
Anomaly detection
Deviations flagged with no alert rules to write or maintain.
Capacity forecasting
Disk-full dates named early, so growth becomes a plan instead of a page.
PromQL explorer
Instant and range queries with presets, for when you want the raw series.
Alert timeline
Severity history in one strip, so incident reviews stop being archaeology.
Scrape target health
Which exporters are up, which are down, and which are flapping.
Governed, like everything on Quox
Reads everything. Changes nothing.
Plugins on Quox are not free-roaming browser extensions. Prometheus Intelligence runs inside the platform’s governance layer, the same one your agents answer to.
- Read-only surface: nothing it does can write to or reconfigure your Prometheus
- Every query routes through the collector proxy, so Prometheus stays on the private network
- Activity is logged through the audit pipeline and witnessed on the WARD chain
SURFACE READ-ONLY fleet heatmap every host, load as heat anomaly detection no alert rules needed capacity forecast disk-full dates, early promql explorer instant + range queries alert timeline severity history TARGETS node-exporter (14) up pve-cluster (3) up blackbox-probes (9) 1 down CONFIG route: /monitoring/prometheus requires: Prometheus server
requires QuoxCORE 1.0+ · requires a Prometheus server · licence: commercial · unlock: licence key in Settings, the Metrics tab appears in the NOC panel