​
What is Atlas?

Atlas is an AI agent orchestration platform that enables engineers to create workspaces where humans collaborate seamlessly with specialized, autonomous agents. Think of it as an operating system for AI agents - providing the infrastructure, supervision, and coordination needed to make AI agents work together effectively.

​
Key Concepts

​
🏠 Workspaces

Self-contained environments where agents live and work. Each workspace has its own configuration, agents, and workflows.

​
🤖 Agents

Autonomous AI entities that perform specific tasks. Atlas supports LLM agents (like Claude), system agents, and custom agents.

​
📡 Signals

Events that trigger workflows. Signals can come from CLI commands, webhooks, schedules, or other sources.

​
💼 Jobs

Workflows that define how agents work together to accomplish tasks. Jobs are triggered by signals and executed by agents.

​
🧠 Memory

Hierarchical memory system that allows agents and sessions to learn from past experiences and maintain context.

​
Why Atlas?

For Individual Developers:
  • Automate repetitive tasks with AI agents
  • Build personal AI assistants and tools
  • Experiment with multi-agent workflows
  • Learn AI orchestration patterns
For Teams:
  • Standardize AI agent usage across projects
  • Share agents and workflows
  • Maintain governance and security
  • Scale AI capabilities organizationally

​
Architecture at a Glance

Atlas uses a daemon-based architecture:
  1. Atlas Daemon - Central service managing all workspaces
  2. Workspace Runtime - Isolated environments for agent execution
  3. Supervisors - Intelligent orchestrators coordinating agent work
  4. Worker Isolation - Each agent runs in secure Deno Web Workers

​
Next Steps

Ready to dive in? Head to our Quickstart Guide to get Atlas up and running in minutes.