Slurm-web gives HPC cluster operators and users a browser dashboard for watching jobs, nodes, and GPUs across a Slurm cluster instead of living in squeue output. The overview page charts resource and job status with interactive graphs, so a glance tells you whether the machine is busy, idle, or quietly failing. The jobs view filters and sorts instantly, updates live as jobs change state, and uses colored badges so pending, running, and failed work sorts itself out visually at any scale. Node views go deeper: advanced visualizations show node status across racking topology, GPU utilization is tracked alongside CPU resources, and problems surface as highlighted issues rather than a line in a log file. Account trees, quality-of-service settings, and advanced reservations get intuitive graphical views that make fair-share politics legible to everyone involved. Prometheus integration collects and charts timeseries metrics for longer-term capacity questions. Access is enterprise-ready: LDAP and Active Directory authentication, OpenID Connect single sign-on through providers like Keycloak or Authentik, and granular role-based permissions decide who sees which clusters. Several clusters can be monitored from one deployment, the interface supports dark mode and custom branding with your own colors and logos, caching keeps the UI responsive against busy controllers, and integrated help explains each view in place. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.