core concepts

Understanding how twig works helps you get the most out of it. This section covers the fundamental concepts behind twig's approach to development.

the twig philosophy

Twig is built on three core ideas:

  1. agents, not assistants - AI should be able to take action, not just provide suggestions
  2. context from analytics - your users' behavior is the best guide for what to build
  3. isolation for safety - work on multiple tasks without fear of conflicts

learning progressively

Start with the basics and gradually explore advanced features:

  1. understand the concepts - read this section to learn how twig works
  2. use the agent - practice prompting and reviewing changes
  3. add rules - encode patterns as you discover them
  4. explore advanced features - try MCP integration when needed

You don't need to understand everything at once. Learn as you go.

key concepts

agentic development

Traditional AI coding tools suggest code that you copy and paste. Twig's agent can read your codebase, run commands, and make changes directly. You stay in control by approving changes before they're applied.

learn more about agentic development

workspaces

Twig uses Git worktrees to create isolated workspaces for each task. This means you can work on multiple features or bug fixes simultaneously without them interfering with each other.

learn more about workspaces

posthog integration

By connecting to PostHog, twig can access your product analytics. This means the agent can reference real user data when helping you make development decisions.

learn more about posthog integration

glossary

New to twig? The glossary defines key terms you'll encounter:

view glossary

next steps