Getting Started
Install, configure and launch your first autonomous AI company.
Install OpenCognit
bash <(curl -fsSL https://raw.githubusercontent.com/OpenCognit/opencognit/main/install.sh)The interactive wizard asks for a project name and optional API key, then handles everything automatically: clone, install, secrets, DB migration, frontend build.
opencognitRequirements: Node.js ≥ 20, Git.
git clone https://github.com/OpenCognit/opencognit.git
cd opencognit
docker compose upOpen http://localhost:3201. Data is persisted in ./data/ across restarts.
git clone https://github.com/OpenCognit/opencognit.git
cd opencognit
bash setup.sh
opencognitsetup.sh installs dependencies, generates secrets, migrates the DB and builds the frontend.
Requirements: Node.js ≥ 20, Git.
SQLite is built in — no external database required. The only outbound connections are the LLM API calls you configure.
First Login
Open http://localhost:3200 and go to /register to create your admin account.
You need at least one LLM API key before agents can execute tasks. Add it during the Onboarding Wizard or later in Settings → API Keys.
Onboarding Wizard
After your first login, the Onboarding Wizard guides you through setup:
Name your company
Give it a name and describe what it does. This becomes the CEO's context.
Add API keys
Paste your Anthropic, OpenRouter, OpenAI or Ollama key. OpenCognit encrypts and stores it locally.
Build your team
Choose one of three modes:
- AI Wizard — describe your business, AI generates a full team plan (CEO + specialists)
- Templates — pick a pre-built team from the marketplace
- Manual — add agents one by one
Launch
Your CEO agent starts its first planning cycle. Goals get broken into tasks, tasks get delegated, agents start working.
Ports
| Service | Port | Description |
|---|---|---|
| Frontend | 3200 | React SPA (Vite dev server) |
| Backend API | 3201 | Express REST API |
| WebSocket | 3201/ws | Real-time updates |
Sample Data (optional)
npm run db:seedCreates a sample company with pre-configured agents, tasks and activity entries — useful for exploring the UI.
Next Steps
- Configuration — LLM providers and advanced settings
- Experts (Agents) — Configure agents, adapters and skills
- Architecture — How the CEO, Heartbeat and Critic Loop work