Find your path
You don’t need to read everything. ArgentOS is deep, but most people only need a handful of pages to get productive. Pick the tier that matches where you are, or jump straight to the use case that fits.By experience level
Getting Started — ~30 minutes
Just landed? Start here.Install ArgentOS
One command, full stack. See Installation.
Onboarding
Set your name, personality, and first model. See Onboarding.
Connect a channel
Wire up Telegram, Discord, or Slack. See Channels.
Power User — ~2 hours
You’ve talked to your agent. Now make it useful.Memory system
How your agent remembers (and forgets).
Model routing
Automatic tier selection so you don’t overpay.
Skills
Teach your agent new capabilities with markdown files.
Marketplace
Install community skills and connectors.
Voice and TTS
Give your agent a voice in the dashboard.
Dashboard deep dive
Tasks, projects, canvas, config panel.
Builder / Operator — ~4 hours
Running agents for a team or building on the platform.Intent governance
Three-tier policy engine (Global > Department > Agent).
Execution worker
Autonomous background task processing.
Knowledge library
RAG with collection-level ACL.
Multi-agent families
Shared memory, inter-agent streams.
Plugins
Integrate external services.
Contemplation
The always-on thinking loop.
Exec approvals
Gated review and sign-off workflows.
By use case
Personal AI Assistant
You want an agent that knows you, talks to you, and gets smarter over time. Key docs: Memory > Voice > Channels > ContemplationBusiness Communications Bot
Route conversations across Telegram, Discord, Slack, or WhatsApp with outbound messaging. Key docs: Channels > Outbound messaging > SkillsTask Automation
Let the agent execute work autonomously — scheduling, background processing, project intake. Key docs: Execution worker > SpecForge > TasksMSP / IT Operations
Monitor infrastructure, run operational playbooks, and enforce governance for client environments. Key docs: Skills > Plugins > Intent governance > Exec approvalsDeveloper Platform
Build and publish skills, connectors, and integrations on the AOS Marketplace. Key docs: Marketplace > Skills > PluginsEnterprise Governance
Lock down agent behavior with hierarchical policies, approval gates, and accountability scoring. Key docs: Intent governance > Exec approvals > Heartbeat > Knowledge ACLQuick reference
| Topic | What it does | Docs |
|---|---|---|
| Install | One-command install and first run | Start here |
| Brain | Auto-capture, entity tracking, hybrid search | Memory |
| Talk | Telegram, Discord, Slack, WhatsApp | Channels |
| Think | Always-on background thinking loop | Contemplation |
| Work | Autonomous task processing | Execution worker |
| Route | Complexity scoring, tier selection, cost control | Model router |
| Guard | Hierarchical policy engine | Intent governance |
| Know | RAG with agent-scoped collections | Knowledge |
| Build | Community skills, connectors, packages | Marketplace |
| Ship | Project intake, team lanes, task generation | SpecForge |
Still not sure where to start? Install ArgentOS and just start talking. Your agent will figure out the rest.