A framework, or a SaaS boilerplate?
A boilerplate gives you a fast first day. Intelligo gives you the same first day and a way to take the next fix to auth, billing or credits without merging it by hand.
A boilerplate is a copy.
Intelligo is a dependency, plus source you own.
Both put working sign-up, teams and billing in front of you on the first day. The difference is who owns each file afterwards: with a boilerplate, all of it is yours to maintain. With Intelligo, the engine stays an npm package you upgrade, and only the pages become your source — four kinds of file, four owners.
What changes, topic by topic
| Topic | A typical boilerplate | Intelligo |
|---|---|---|
| What you get | A repository to copy: auth, billing, a dashboard and a landing page, already wired together. | 11 npm packages for the engine, and the pages installed into your app as source with the shadcn CLI. |
| Updates | The kit keeps moving after you fork it. Taking a fix means merging it into code you have changed. | Packages upgrade by version. intelligo upgrade --check says which generated files changed upstream and which you edited, and overwrites nothing. |
| Database changes | Migrations live in your fork; a new one from the kit is yours to reconcile. | intelligo migrate applies the framework's schema by content hash in one transaction; migrate --check gates a deploy. |
| What you edit | Everything — which is also why every later update is a merge. | The pages, through config files and message files they ship; the packages never. Your agent is untouched. |
| Workspaces and roles | Often included; how deep it goes varies by kit. | Multi-tenant workspaces, invitations and role checks on Better-Auth, with every query scoped to a workspace. |
| AI cost and credits | Some kits add a credit counter. Cost per run is usually yours to build. | Every run is admitted, settled or failed once: entitlement checked, worst-case cost reserved, actual usage charged in credits. |
| Audit and operations | Rarely included. | Append-only audit events a Postgres trigger enforces, a Postgres job queue, and an admin console across tenants. |
| Your AI framework | Whatever the kit chose, or none. | Any, used natively. Intelligo has no agent abstraction; the chat route is the AI SDK's. |
| Price and licence | Some are free; most complete kits are a one-time licence. | Free. Apache-2.0, including the pages. |
Boilerplates differ; the left column is what most of them share, not any one kit. The right column is what the docs and the source show.
When a boilerplate is the better choice
- You want a marketing site, blog and docs in the same repository.
- Intelligo ships the product's pages, not a landing page.
- Your stack is not Next.js and Postgres.
- The packages assume both; a kit for your stack will fit better.
- You will never take an update.
- If the fork is the finished product, a snapshot is simpler than a dependency.
- You need a stable 1.0 today.
- Intelligo is at 1.0.0; the API can still change before 1.0.
Run it before you decide
Getting started creates an app, connects Postgres and runs the whole product against a stub model, before you configure an AI provider. The other alternatives — building in-house, an AI platform, prompting a coding agent from zero — are on /why.
Ready to try it? Three commands and the app runs against a stub model.