Scrypted Telemetry Plugin

Hardware monitoring plugin for Scrypted. Collects CPU, GPU, NPU, memory, disk, and network metrics from multiple sources and provides Prometheus-compatible endpoints.

What it does

• Monitors Intel GPU, Intel NPU, NVIDIA GPU, Apple Silicon metrics
• Collects system metrics (CPU temp/usage, memory, disk, network)
• Provides Prometheus endpoints for Grafana dashboards
• Works across multiple Scrypted nodes with cluster sharing

Required exporters

Install these first, then configure the plugin:

Intel systems:

• ghcr.io/bjia56/intel-gpu-exporter
• dmontgomery40/intel-npu-top
• ghcr.io/dmontgomery40/intel-npu-prometheus-exporter

macOS:

• ghcr.io/dmontgomery40/macmon-prometheus-exporter

All systems:

• Prometheus
• Telegraf (optional, for enhanced system metrics)

Installation

1. Install required exporters for your hardware
2. Install this plugin in Scrypted
3. Configure exporter endpoints in plugin settings
4. Access metrics at /endpoint/telemetry/metrics or standalone server port

Screenshots

Grafana dashboard with multi-node monitoring:

Scrypted plugin configuration:

Docker deployments

Pre-configured docker-compose files in /deployments/ for different hardware combinations.

Prometheus config example

scrape_configs:
  - job_name: 'scrypted-telemetry'
    static_configs:
      - targets: ['scrypted-host:9090']

Notes

This was fun to build but I can't scale it for everyone's setup. Take it, run with it, make it easier for users if you want.

Thanks to bjia56 for cosmotop inspiration and koush for Scrypted.