OptimAI Node Operator Toolkit

Real, working tools for operating one or many OptimAI Network nodes from a single control box. The OptimAI CLI (optimai-cli) is the only first-party binary, but it doesn't help when you want to:

• Run a fleet of nodes
• Watchdog a single node and auto-restart on failure
• Pull rewards history and pretty-print it
• Send a Telegram alert when the device disconnects

This repo provides exactly that. It is not a wrapper that hides the official CLI — it calls it via subprocess and adds missing glue.

Files

File	Purpose
monitor.sh	watch the local crawl4ai container and the CLI process; restart on crash; log to syslog
fleet.sh	run the same command on N nodes over SSH (deploy a config update)
rewards.py	parse the CLI's --rewards JSON output, build a CSV, plot a 7-day line
healthcheck.py	long-running daemon that pings the CLI's status and alerts via Telegram on disconnect
Makefile	one-liner targets: install, monitor, rewards
systemd/optimai-monitor.service	systemd unit for the healthcheck

Quick start

make install              # copy scripts to /usr/local/bin and install systemd unit
optimai-monitor --install # tells the CLI to start on boot
make rewards              # prints last 7 days of rewards as a CSV

Requirements

• The official optimai-cli already installed (/usr/local/bin/optimai-cli)
• An OptimAI account already signed in (run optimai-cli signin once)
• Python 3.10+ for the scripts
• Optional: matplotlib for the rewards plot

License

MIT