Quickstart
This guide walks you through installing the Obol Stack, setting up an AI agent, and optionally deploying a blockchain network.
Prerequisites​
- Docker installed and running on your machine.
- macOS or Linux operating system.
- At least 8 GB of RAM available.
Ensure Docker is running before proceeding. You can verify this by running docker info in your terminal.
Step 1: Install Obol Stack​
Run the bootstrap installer:
bash <(curl -s https://stack.obol.org)
The installer will:
- Validate that Docker is running.
- Install the
obolCLI binary and dependencies (kubectl, helm, k3d, helmfile, k9s). - Configure your PATH and add
obol.stackto/etc/hosts. - Offer to start the cluster immediately.
- Default installation
- Specific version
- Development mode
bash <(curl -s https://stack.obol.org)
Files are installed to:
- Config:
~/.config/obol/ - Data:
~/.local/share/obol/ - Binaries:
~/.local/bin/
OBOL_RELEASE=v0.1.0 bash <(curl -s https://stack.obol.org)
git clone https://github.com/ObolNetwork/obol-stack.git
cd obol-stack
OBOL_DEVELOPMENT=true ./obolup.sh
Development mode uses a local .workspace/ directory and runs go run instead of a compiled binary.
Step 2: Start the stack​
obol stack init
obol stack up
The first startup may take a few minutes as Docker pulls the required images.
Step 3: Set up the AI agent​
Initialize an OpenClaw agent instance:
obol agent init
This walks you through choosing a model provider:
- Ollama (local, free) - if Ollama is detected on your machine
- Anthropic or OpenAI - routed through the in-cluster llmspy gateway
Once complete, open the agent dashboard:
obol openclaw dashboard
Each agent is automatically provisioned with an Ethereum signing wallet (via the remote-signer service). The wallet address is displayed during setup.
You can reconfigure the model provider at any time with obol openclaw setup.
Step 4: Deploy a blockchain network (optional)​
Install an Ethereum node on the Hoodi testnet:
obol network install ethereum --network=hoodi
Deploy it to the cluster (replace with your deployment ID):
obol network sync ethereum/<your-id>
Replace the ID with the actual deployment ID shown in your terminal output (e.g., knowing-wahoo).
Check the deployment:
obol kubectl get pods -n ethereum-<your-id>
Step 5: Explore​
# Interactive cluster UI
obol k9s
# View all pods
obol kubectl get pods -A
# Check tunnel status (public URL)
obol tunnel status
# Configure model provider globally
obol model setup
Stopping and cleaning up​
# Stop the cluster (preserves data)
obol stack down
# Restart
obol stack up
# Remove everything including data
obol stack purge -f
obol stack purge -f is irreversible. It removes all cluster data and configuration.
Next steps​
- Installing networks - Deploy different blockchain networks.
- Installing apps - Deploy additional applications.
- FAQ - Common questions and troubleshooting.