What is Sunboard
Agent-native onboarding for SaaS apps — checklists and tours that ship as code.
Sunboard is agent-native onboarding for SaaS apps. An expert AI plans a high-converting activation flow with you, wires it into your product, and ships it as code — versioned with your commits, instead of clicked together in a separate admin UI that drifts from your product.
You don't need to be a growth expert. Sunboard's agent studies your product, then interviews you about your users and the job they're hiring it for — the aha moment, the must-have setup, the segments worth targeting. It turns your answers into an activation strategy and writes the spec. That spec is a real file in your repo, not a setting in someone else's dashboard.
What you can build
Checklists
A guided list of the steps that lead a new user to their first real win.
Product tours
Short, focused walkthroughs that point at the real buttons and routes in your app.
Every step anchors to real UI through data-sunboard-id selectors, and
completes when a real product event fires via useSunboard().track(). Because
your agent already knows your codebase, the anchors land on the right elements.
How it fits together
The sunboard CLI
Local commands your agent runs to author, preview, and deploy. Every command supports --json for AI-friendly output.
@sunboard/react
The SDK that renders the checklist, tours, and hotspots in your app. Installed by sunboard init.
Sunboard (hosted)
Where you promote candidate versions, bind audience segments, and read activation analytics.
Your AI agent
The reasoning layer — Claude Code, Codex, or opencode — driving the whole workflow through first-party skills.
The workflow
You drive everything from your AI agent. Each skill maps to a stage in a loop you can repeat as your product grows:
- Spec — interview-driven planning that drafts the onboarding experience as YAML in your repo.
- Implement — wire the
data-sunboard-idselectors andtrack()calls into your real product. - Preview — boot your dev server with the spec served against your real app, before anyone sees it.
- Deploy — upload the spec as an immutable candidate version with its git metadata. It does not go live automatically.
- Promote — review the candidate in Sunboard, target the right segments, and publish. Old versions stay available, so rollback is one click.
- Analyze — read the real funnel, find where users drop off, and ship a concrete fix. Then measure again.
Reviewed before it goes live
Deploying only creates a candidate. Publishing is a separate, authenticated step in Sunboard — never something that happens from your local machine. A public runtime key can read onboarding but never activate it.
How it's different
Tools like Pendo, Appcues, or Userpilot build onboarding in a separate visual dashboard that drifts from your product. Sunboard authors it as code in your repo, so your onboarding gets the same history, review, and blame as the rest of your product — and your AI agent does the wiring against the real codebase. It's a natural fit for teams shipping with AI coding agents, including AI-built and vibe-coded SaaS apps.