cluster-job-monitor¶
A read-only terminal dashboard that shows your SLURM jobs across several
clusters and your desktop in one view. It SSHes into each host, runs
squeue --me, and renders a live, colour-coded overview.
Get started in the Quick start :material-arrow-right:{ .md-button .md-button--primary }
Read-only by design¶
The only commands ever run are squeue and sinfo (SLURM hosts) and
nvidia-smi + ps (non-SLURM GPU hosts). There is no code path that can
cancel or submit jobs. It uses your existing SSH config and keys — nothing new
is exposed, no server, no stored secrets.
Features¶
- One view across hosts — every cluster, plus your local desktop, in a single live TUI.
- For coding agents — a one-shot
--overview(CLI or MCP) returns, per cluster and partition, how many CPUs/GPUs are free — broken down by GPU type, with the largest free block on a single node — plus your queued/running jobs and an approximate queueing time. - Non-SLURM GPU boxes — point it at a workstation with
"scheduler": "gpu"and it shows one row per GPU process vianvidia-smi. - No-cluster demo — try the whole thing with synthetic data using
--demo. - Keyboard-driven — collapse/expand clusters, filter by state, cluster, partition, or job name.
Where to next¶
- Quick start — install and run the demo in under a minute.
- Configuration — point it at your real clusters.
- Agent overview — the capacity JSON for agents.
- MCP — expose the overview to Claude Code / Cursor.
- Development — tests, coverage, CI.