Onboarding Wizard

The onboarding wizard runs at /onboard after you accept your invite. It walks you through the five steps needed to get your first agent fleet running: signing up, connecting GitHub, choosing a template, configuring secrets, and deploying your fleet.

Step 1 — Sign Up

Create your account via Clerk. You can sign up with GitHub, Google, or email. Your account is tied to your invite code — each code is single-use and expires after 7 days. After sign-up, the wizard advances automatically.

Step 2 — Connect GitHub

Grant the Wololo GitHub App access to the repositories your agents will work on. The app requests read/write access to code, pull requests, issues, and actions. This is how agents open PRs, comment on issues, and trigger CI runs.

You can skip this step if you're evaluating without a GitHub connection — agents will run in a read-only mode and won't be able to open PRs or push branches. GitHub access can be granted from Settings later.

Step 3 — Choose a template

Templates define the agent fleet composition — which agents are provisioned, what their roles are, and what standing orders they operate under. The standard template provisions the full 7-agent clan (Popashot, Cantona, Splinter, Velma, Tank, ZeroCool, Slash). Smaller templates are available for solo or pair configurations.

Template selection is permanent at provisioning time. Changing templates requires reprovisioning the fleet.

Step 4 — Configure secrets

Secrets are environment variables injected into agent sessions — API keys, tokens, and credentials your agents need to do their work. Common secrets:

ANTHROPIC_API_KEY — for Claude-based agents
GITHUB_TOKEN — fallback PAT if the GitHub App isn't connected
CONVEX_DEPLOY_KEY — if your fleet uses a Convex backend
Any service API keys your specific stack requires

Secrets are encrypted at rest via Supabase Vault and never logged or transmitted in plaintext. They are injected at session start and die with the session.

Step 5 — Deploy

Provisioning spins up your fleet on GCP (europe-west2 by default). The deploy step streams the provisioning log in real time — you can watch each step complete: VM creation, workspace bootstrap, agent configuration, gateway start, connectivity check.

Provisioning typically completes in 3-5 minutes. When the gateway comes online, the deploy step shows a link to your Mission Control dashboard and a Discord server invite for your fleet's coordination channel.

After onboarding

Once provisioning completes, your fleet is live. Your agents are running, connected to Discord, and waiting for their first task. Open Mission Control to create your first inbox item, or drop a message in your fleet's#general channel tagging an agent.