Getting Started
Get AMOF running and your first multi-repo workspace bootstrapped in under 5 minutes.
Prerequisites
| Tool | Required | Purpose |
|---|---|---|
| Python 3.10+ | Yes | AMOF CLI and orchestrator |
| Git 2.30+ | Yes | Worktree support |
| Docker | For local clusters | k3d provisioner |
| Terraform | For cloud provisioners | AWS, GCP, Azure, DO, Hetzner |
| kubectl | For Kubernetes ops | Cluster interaction |
| Helm 3 | For chart deployments | Kubernetes package management |
Optional but recommended: Cursor or VS Code with the Dev Containers extension, and an LLM API key (Anthropic, OpenAI, or OpenRouter).
amof checkInstallation
Option 1: DevContainer (Recommended)
The DevContainer includes all tools pre-installed. The container mounts ~/.aws and ~/.ssh automatically for credential access.
git clone https://github.com/your-org/amof.git
cd amofOpen in Cursor or VS Code and click "Reopen in Container" when prompted.
Option 2: Local Installation
git clone https://github.com/your-org/amof.git
cd amof
# Install Python dependencies
pip install -r requirements.txt
# Add the amof alias to your shell
source scripts/amof-aliases.sh
# Verify installation
amof checkConfigure Your Environment
Create a .env file in the project root with your API keys and credentials:
# LLM Provider (pick one or more)
ANTHROPIC_API_KEY=sk-ant-...
OPENAI_API_KEY=sk-...
OPENROUTER_API_KEY=sk-or-...
# Git credentials (for private repos)
GIT_TOKEN=ghp_...
# Optional: Jira / Confluence
ATLASSIAN_USER=your.email@company.com
ATLASSIAN_TOKEN=...Set your default ecosystem in .amof/agent.yaml:
default_ecosystem: my-projectQuickstart: Local Development with k3d
The fastest way to experience AMOF -- no cloud credentials needed.
1. List Available Ecosystems
amof ecosystem list2. Bootstrap a Workspace
amof -e supabase-platform installThis creates a Git worktree at worktrees/supabase-platform/, clones all repositories defined in the manifest, generates repo profiles, and builds the codebase index.
3. Check Status
amof -e supabase-platform statusREPO BRANCH COMMIT MODE STATUS
supabase-kubernetes workspace/supa... a1b2c3d RW OK
supabase-config workspace/supa... e4f5g6h RW OK
shared-infra main i7j8k9l RO OK4. Spin Up Local Infrastructure
amof -e supabase-platform spin deployThis creates a local k3d Kubernetes cluster with load balancer ports mapped to your host.
5. Start Working on a Feature
# Create feature branches in all writable repos
amof ticket start add-auth-flow
# Check the new branches
amof status6. Run the AI Agent
# Interactive mode
amof agent
# Single-shot task
amof agent "Add a healthcheck endpoint to the backend service"
# Read-only analysis
amof agent --plan "Explain the authentication flow across all repos"7. Push and Clean Up
# Push all branches to remotes
amof push
# When finished, archive the workspace
amof archive -m "Auth flow complete"
# Tear down infrastructure
amof -e supabase-platform spin destroyQuickstart: Cloud Deployment with AWS
# 1. Bootstrap
amof -e aws-boilerplate install
# 2. Configure Terraform Variables
# Create ecosystems/aws-boilerplate/spin.tfvars
# 3. Deploy Cloud Infrastructure
export AWS_ACCESS_KEY_ID="..."
export AWS_SECRET_ACCESS_KEY="..."
amof -e aws-boilerplate spin deploy
# 4. Work, Push, Destroy
amof ticket start feature-login
amof agent "Implement OAuth2 login in the frontend"
amof push
amof -e aws-boilerplate spin destroyUsing the Control Plane UI
AMOF includes a web-based control plane for visual management.
amof server startThis launches the Backend API at http://localhost:8000 (Swagger at /docs) and the Frontend UI at http://localhost:3000 (Next.js dashboard).
- Ecosystem Grid -- See all ecosystems with status, agent count, and last indexed time
- Orchestrator Log -- Real-time SSE-streamed terminal output with ANSI colors
- Agent Chat -- Interactive AI chat with ask/plan/execute modes
- Task Runs -- Monitor active and completed agent runs
- Release Management -- Bump versions, promote pre-releases, view audit log
What's Next?
| Topic | Link |
|---|---|
| Understand the system design | Architecture |
| Configure ecosystems and provisioners | Ecosystems & Provisioning |
| Master the AI agent system | AI Orchestration |
| Explore the API and UI | Control Plane |
| Learn every CLI command | CLI Reference |
| Configure AMOF | Configuration |