User Guide

Walkthrough of every NOBA feature, from first login to advanced operations.

Getting Started

After installing NOBA, retrieve your admin password from the journal, open the web UI at http://your-host:8080, and log in. The first-run wizard guides you through adding your first integration.

Roles

Dashboard

The main dashboard shows all monitored targets as cards with live metrics via Server-Sent Events. Cards show sparkline charts, health bars, and quick-action buttons. Press g to collapse all cards to headers for a quick overview.

Custom Dashboards

Create custom dashboards with widgets for any metric, chart, or integration data. Widgets include time-series charts, gauges, counters, and markdown panels. Dashboards can be shared with other users.

Remote Agents

Deploy agents from Agents → Deploy via SSH or one-liner. Each agent is a zero-dependency Python zipapp that reads /proc directly and connects over encrypted WebSocket. Capabilities include 42 command types across 3 risk tiers, remote desktop (Wayland/X11/Windows/macOS), embedded terminal (xterm.js), file transfer (50 MB, SHA-256 verification), and self-update.

Integrations

50+ integrations across infrastructure, media, network, IoT, DevOps, and auth are listed from Settings → Integrations. Connector setup, test buttons, and wizard steps appear where that connector exposes them.

Automations & Workflows

10 automation types including scripts, webhooks, service restarts, workflows, conditions, delays, notifications, HTTP calls, agent commands, and remediation actions. The visual workflow builder supports conditional branching, approval gates, cron scheduling, and maintenance windows.

Self-Healing Pipeline

6-layer healing workflow: correlation, dependency analysis, planning, execution, verification, and learning. Use observation, dry-run, notify, or approval-gated modes while validating each wired remediation action and escalation chain against your own environment.

AI / LLM Intelligence

Configure under Settings → Integrations → AI / LLM. Supports Anthropic (Claude), OpenAI (GPT-4o), Ollama (local), and any OpenAI-compatible endpoint. The AI receives live infrastructure context with every request. Completely optional — off by default.

Monitoring & SLA

SLA dashboards show 7d/30d/90d uptime percentages per agent/service. Composite alert rules support AND/OR conditions with escalation policies. Endpoint monitoring for HTTP/HTTPS checks with cert expiry alerts.

Health Score

Per-agent health scoring using weighted composite: uptime 40%, latency 25%, error rate 20%, headroom 15%. Grades from A (excellent) to F (critical).

Security Posture

Per-agent security scoring with findings dashboard and config baseline drift detection. Agent-based scanning with categories including TLS, firewall rules, SSH config, sudo permissions, and more.

Status Page

Public status page with component groups, 90-day uptime charts, incident lifecycle management. No authentication required — share with your users.

Multi-Site Operations

Monitor and heal across physical sites. Cross-site endpoint checking, HMAC-SHA256 webhook chains, agent multi-homing, and independent failure domains. Each site operates autonomously with cross-site endpoint verification.

Infrastructure Export

Export your infrastructure configuration as Terraform, Ansible, or CloudFormation templates for infrastructure-as-code workflows.

Updating

Self-update from Settings → General → Self-Update. The updater pulls changes, verifies integrity, installs dependencies, and restarts — with automatic rollback on failure.