Fabbro · development companion

From idea to production. One opinionated path.

Fabbro already made the hard architecture calls, so you — or your AI agent — build a secure, coherent, production-ready app instead of wiring one. Ship on a monolith with SSR and SQLite; graduate to Postgres, Redis, and a background queue without switching frameworks.

Get started Open the board
Opinionated by design

Where Fabbro goes against the grain

Strong defaults are the point. Each of these is a deliberate call the framework enforces — not a limitation, a position.

The market says: microservices + SPA

A monolith with SSR, on purpose.

One deployable app, server-rendered, frontend coupled by default. Fewer moving parts, one flow to read and debug — the fastest way from an idea to something running.

The market says: a dedicated database

SQLite in WAL, until you prove otherwise.

It covers most real-world workloads with zero setup. The moment your load says you need more, Fabbro names the threshold and hands you Postgres — same framework, no rewrite.

The market says: AI writes spaghetti

AI development can be safe and coherent.

Anchors, typed errors, no hardcoded config or secrets, a doctor that verifies wiring. The whole framework is built to make AI-generated code reviewable and correct by construction.

Scale when you do

The same framework, one level deeper

Growth is a mode you switch on, not a tool you swap. Start simple; turn these on when the workload asks.

Postgres

Graduate from SQLite to a dedicated database when concurrency and volume call for it.

planned

Redis

Move rate-limiting and shared state off in-process memory to a global store across workers.

planned

ARQ queue

Swap the SQLite job queue for a real background worker when the volume of async work grows.

planned

Backend API

Run Fabbro headless as an API-only backend, with your own API surface and no server-rendered UI.

planned

Your development partner, not just a scaffolder.

One place to centralize decisions, artifacts, documentation, and planning — a 360° view of the project, living right next to the code it describes. No context scattered across five other tools.

Playbooks

Interactive, copy-ready guides from a fresh project to a running app.

Open

Board

Plan modules, UIs, and tasks with checklists, deadlines, and notes.

Open

Skills

Context files to load into your AI agent so it knows this project.

Browse

Prompts

Ready-made prompts to steer an AI development session the Fabbro way.

Browse
Playbooks

Guides that run where you build

Each playbook is a focused walkthrough — getting started, anatomy of a project, building a module, the registry, the design system, developing with AI. Commands are one click to copy, and the checklists remember what you've done across restarts. It's the docs, but interactive and next to your code instead of in a browser tab.

Board

Project planning, finally in one place

A kanban built for how devs actually work: pre-made cards for modules, UIs, API integrations, and setup — each with the Fabbro checklist already filled in. Two deadlines (project and card), a notes drawer for tech debt and client changes, and light gamification that turns "did I finish?" into a number. One board per project, because each project already has its own _dev/.

Skills

Context library for your AI agent

Load these into Claude Code, Codex, or any coding agent so it knows this project's architecture. Add your own — they're just .md files.

loading…

Prompts

Prompt library

Ready-made prompts that already phrase a task in Fabbro's terms. Add your own to reuse across sessions.

loading…