atlas-os
v1.7.5
Published
Atlas·OS — Autonomous Teams · Lifecycle · Agents · Skills — Orchestration System for the terminal.
Downloads
1,754
Maintainers
Readme
ATLAS·OS
Autonomous Teams · Lifecycle · Agents · Skills — Orchestration System
A multi-agent, hook-driven, model-agnostic engineering OS for the terminal. Hand it a vague idea. Get back a planned, built, tested, committed feature — with a Greek pantheon of specialist agents doing the work.
npx atlas-os@latestWorks on macOS, Linux, and Windows through WSL2. Bring Anthropic Claude, OpenAI GPT, OpenRouter, OpenCode Zen, OpenCode Go, or run local Llama / Qwen / DeepSeek models through Ollama, LM Studio, vLLM, or llama.cpp.

Install
macOS / Linux
npx atlas-os@latest
# Or install globally
npm install -g atlas-os
atlasAtlas checks npm for newer atlas-os releases when the TUI opens. If your
global install is behind, it shows a short dismissible update notice with:
npm install -g atlas-os@latestWindows
WSL2 is recommended:
wsl --install -d UbuntuThen inside Ubuntu:
sudo apt update && sudo apt install -y nodejs npm
npm install -g atlas-os
atlasNative Windows can run the core CLI, but the shell tool expects POSIX commands.
VS Code
atlas vscode-setup
# then reload VS CodeUse --dry-run to preview the settings change.
Providers
Open the TUI and use /config to add a provider key:
atlasEnvironment variables also work:
export OPENROUTER_API_KEY=sk-or-...
export ANTHROPIC_API_KEY=sk-ant-...
export OPENAI_API_KEY=sk-...
export OPENCODE_ZEN_API_KEY=oc-...
export OPENCODE_GO_API_KEY=oc-...Hosted model cost posture is configurable from /config -> Atlas power mode.
| Mode | Cost Estimate | Pros | Cons | |---|---|---|---| | Atlas Power Full | roughly 100k-250k input tokens on heavy turns before cache; cache-capable models make repeat turns cheaper | maximum Atlas context, tools, MCP, hooks, and predictable behavior | no-cache models rebill the full prefix each message | | Atlas Smart Mode | roughly 20k-80k input tokens on normal hosted turns; complex turns can still pay Full Atlas costs | cost-aware default for daily hosted work | very complex work may still need the full prompt/tool surface |
The active hosted mode is visible in the TUI top bar as ATLAS POWER in bright
red or ATLAS SMART in bright green.
The model picker writes cache support beside provider-pulled models as
cache yes, cache unknown, or cache no so cheaper cache-capable models are
easy to spot. OpenRouter and OpenCode rows use live /models cache-pricing
fields, and /models includes a search field for filtering long provider
catalogs.
Saved sessions are managed from /sessions: resume, rename, start fresh,
delete one, select several for deletion, or delete all with confirmation.
Local models work through Ollama, LM Studio, vLLM, or any local
OpenAI-compatible /v1 server. Atlas auto-detects Ollama at
http://localhost:11434/v1; use /config -> Local models to choose Lite,
Hybrid, or Full Atlas mode.
ollama pull qwen2.5-coder:1.5b
ollama pull qwen2.5-coder:7b| Mode | Requirements | Best For | Tradeoff | |---|---|---|---| | Lite | CPU ok, 4-8 GB RAM, 1.5B-7B models | quick local chat | no model-driven tools | | Hybrid | 8-12 GB VRAM or strong CPU, 7B-14B models | local coding with core dev tools | limited tool set | | Full Atlas | 24 GB+ VRAM or hosted server, 30B-70B+ models | full Atlas prompt, tools, MCP, and hooks | largest payload |
Bootstrap once:
atlas initNew launches start fresh. Use /sessions or /resume <id> to reopen old
transcripts manually.
What It Does
ATLAS·OS is a multi-agent, spec-driven coding system for the terminal. You describe the goal; the orchestrator routes work across specialist agents for planning, architecture, implementation, QA, and release.
Under the hood: typed tool contracts, lifecycle hook guardrails, skills, checklists, MCP support, sessions, and provider/model routing.
Why It Is Different
| Capability | ATLAS·OS | Claude Code | OpenCode | Gemini CLI | Kilo Code | |---|---|---|---|---|---| | Provider choice | Anthropic · OpenAI · OpenRouter · OpenCode Zen/Go | Claude-focused | Provider-agnostic | Gemini-focused | Kilo router | | Multi-agent orchestration | Built-in Greek pantheon | Agent teams + subagents | Build / Plan + subagent | Subagents | Modes | | Spec-driven pipeline | Built in | Bring your own | Bring your own | Bring your own | Bring your own | | Lifecycle hooks | Typed TypeScript hooks | Hook system | Plugins / MCP | Hooks | Plugins / MCP | | MCP servers | Built in | Built in | Built in | Built in | Built in | | Terminal-first | Full-screen TUI | Terminal | Terminal | Terminal | VS Code-first | | License | MIT | Proprietary | MIT | Apache-2.0 | MIT |
Atlas's edge is the SDD pipeline, specialist agents, typed hooks/tools, MCP integration, and release workflow shipped as one terminal system.
