Monitoring
Real-time system health, 25+ dashboard widgets, and AI-powered diagnostics.
Talome gives you real-time visibility into your server's health — CPU, memory, disk, network, GPU, containers, and every integrated app. The dashboard shows it at a glance. The AI diagnoses it on demand.
Screenshot: Dashboard home showing system health widgets: CPU, memory, disk, network, and container status
Dashboard Widgets
The dashboard home page is built from customizable widgets. Talome ships with 25+ built-in widgets organized by category.
System Widgets
| Widget | What It Shows |
|---|---|
| CPU | Real-time CPU usage graph with per-core breakdown |
| Memory | RAM and swap usage with historical trend |
| Disk | Per-mount usage, free space, and growth rate |
| Network | Throughput in/out with real-time graph |
| GPU | GPU utilization, memory, temperature, and encoder/decoder usage |
| System Health | Overall health score combining CPU, memory, disk, and container status |
| System Info | OS, kernel, uptime, Docker version, and hardware details |
| System Status | Quick traffic-light status for all monitored subsystems |
| Storage Mounts | All mounted drives with usage bars and mount points |
App and Service Widgets
| Widget | What It Shows |
|---|---|
| Services | Grid of installed apps with health status indicators |
| Arr Status | Sonarr/Radarr queue, upcoming, and missing counts |
| Overseerr Requests | Pending and approved media requests |
| Ollama Status | Loaded models, VRAM usage, and inference stats |
| Active Downloads | qBittorrent download queue with progress bars |
| Audiobooks | Audiobookshelf listening progress and library stats |
| Backup Status | Last backup time and success/failure for each app |
Media Widgets
| Widget | What It Shows |
|---|---|
| Media Calendar | Upcoming TV episodes and movie releases from Sonarr/Radarr |
| Optimization | Media transcoding queue status and space savings |
Utility Widgets
| Widget | What It Shows |
|---|---|
| Quick Actions | Configurable shortcut buttons for common tasks |
| Activity | Recent actions and events across all apps |
| Digest | AI-generated summary of what happened since your last visit |
| Launcher | Quick-launch grid for installed app web UIs |
| List | Custom list widget — bookmarks, notes, or any structured data |
| Stat Tile | Single-stat display for any metric (CPU temp, disk free, etc.) |
| Divider | Visual separator between widget groups |
| Declarative | Custom widget defined by a JSON schema — for advanced users |
| Bento Gallery | Photo grid widget pulling from your media library |
Widget Customization
Drag-and-Drop Layout
Widgets can be rearranged by dragging on the dashboard. Resize them by pulling the corner handle. The layout is saved per-user.
AI-Driven Arrangement
Tell the AI how you want your dashboard:
The create_widget_manifest and update_widget_manifest tools manage the widget layout. The list_widgets tool shows what's currently on the dashboard.
Creating Custom Widgets
The AI can create custom widgets through conversation using the create_widget_manifest tool. Describe what you want to see, and the AI generates a declarative widget definition that appears on your dashboard immediately:
You can also create entirely custom widget types using the declarative widget system, which accepts a JSON schema defining data sources and display format.
System Health Tools
The AI has direct access to system metrics — no widgets required:
| Tool | What It Returns |
|---|---|
get_system_stats | CPU, memory, disk, and network usage in one call |
get_system_health | Overall health assessment with per-subsystem scores |
get_disk_usage | Detailed per-mount disk usage with inodes |
get_gpu_status | GPU utilization, memory, temperature, running processes |
get_smart_status | S.M.A.R.T. health data for all drives |
get_metrics_history | Historical metrics over a time range for trend analysis |
Conversation Examples
The AI calls get_system_health and gives you a plain-English summary: CPU at 12%, memory at 4.2GB/16GB, all disks healthy, 23 containers running with 22 healthy.
The AI calls get_smart_status, reads the reallocated sector count, power-on hours, and temperature, and tells you whether to worry.
The AI calls get_metrics_history with a 24-hour window and summarizes the peaks, averages, and any anomalies.
Container Monitoring
Every running container is monitored for health, resource usage, and logs.
| Tool | What It Does |
|---|---|
get_container_stats | CPU, memory, network, and disk I/O for a specific container |
check_service_health | Runs the container's healthcheck and reports pass/fail with details |
analyze_service_health | Deep analysis — checks logs, resource trends, and connectivity |
search_container_logs | Full-text search across a container's log output |
get_container_logs | Returns the last N lines of a container's logs |
Conversation Examples
The AI checks all containers and ranks them by CPU and memory usage.
The AI calls analyze_service_health, reads the logs for OOM kills or crash traces, checks memory limits, and reports the root cause.
The AI calls list_containers, filters for unhealthy ones, and runs diagnose_app on each.
Storage Monitoring
Disk space is the most common home server emergency. Talome helps you stay ahead of it.
| Tool | What It Returns |
|---|---|
get_storage_breakdown | Space usage by category — apps, media, backups, Docker images, system |
get_reclaimable_space | Unused Docker images, build cache, stopped containers, and orphaned volumes |
cleanup_docker | Removes reclaimable resources with a dry-run option |
get_disk_usage | Per-mount usage with growth trends |
Conversation Examples
The AI calls get_storage_breakdown and shows you a categorized breakdown — 800GB media, 12GB Docker images, 3GB app data, 2GB backups.
The AI calls get_reclaimable_space to find unused images, dangling volumes, and build cache. Then offers to run cleanup_docker with your confirmation.
The AI checks get_disk_usage trends, calculates the growth rate from get_metrics_history, and projects when you'll hit capacity.
Automated Monitoring
Combine monitoring tools with automations for proactive alerts:
The automation engine can run any monitoring tool on a cron schedule and trigger notifications when conditions are met.
Monitoring Automation Ideas
| Automation | Schedule | What It Does |
|---|---|---|
| Disk space alert | Every hour | Checks disk usage, notifies if > 90% |
| Container health check | Every 5 minutes | Checks all containers, notifies on unhealthy |
| Daily digest | Daily 8:00 AM | Summarizes overnight health, downloads, and issues |
| SMART check | Weekly | Reads drive S.M.A.R.T. data, warns on degradation |
| Docker cleanup | Weekly | Removes unused images and build cache |
| Resource hog alert | Every 15 minutes | Notifies if any container uses > 80% CPU for sustained period |