Why Habits?

Build backend logic (agents, automations, SaaS) using pre-built, tested building blocks instead of custom code, then use AI to generate the frontend.

The Problem

Building logic requires connecting to APIs, handling errors/retries, securing vulnerabilities, and tracking execution. When you write custom code, all these responsibilities are on you.

Current tools run in a centralized way, with almost no way to audit the runtime or decentralize specific workflows. Making it impossible to pack specific workflows to run on serverless functions or as standalone apps.

The Solution

Habits provides pre-built nodes that handle the hard parts:

nodes:
  - id: send-email
    type: activepieces
    data:
      module: "@ha-bits/bit-database"
      operation: insert
      label: Store in DB
      params:
        collection: survey-responses

Key Benefits

Benefit	What it means
Pre-built integrations	200+ connectors for common services (APIs, databases, AI, etc.), we are gradually open-sourcing those and releasing to public NPM
Less code	Configure what you want, not how to do it
Easy to audit	YAML files show exactly what happens in each workflow
Mix and match	Combine nodes from different platforms (Habits, Activepieces, n8n) in one workflow

Three Ways to Build

Approach	When to use	Example
Pre-built nodes	Standard integrations	OpenAI, Postgres, SendGrid, etc
Declarative config	Simple HTTP calls, data transforms	REST API requests
Custom scripts	Unique business logic	Deno/TypeScript inline code

Open Source

Habits is Apache 2.0 licensed.

Node Licensing

When using nodes from different ecosystems, check their licenses:

• Bits & ActivePieces nodes → MIT (fully permissive)
• n8n nodes → SUL (some commercial restrictions)

Why not bundle n8n and Activepieces?

Habits initially aimed to be a runtime (The cortex part) that runs n8n and Activepieces nodes for SaaS backends. However, n8n's source-available licensing created constraints, and Activepieces' single-trigger model didn't fit our multi-trigger, out-of-order execution needs. We kept some compatibility with both ecosystems while introducing our own Bits for flexibility.

Why Not Vibecode Everything?

AI-generated code is fast but often vulnerable. Habits takes a pragmatic approach:

Layered Security Model

1. Frontend – Vibecode freely (browser-sandboxed, XSS mostly mitigated)
2. Backend Bits – Vibecode with guardrails (SAST, SCA, partial DAST before packaging)
3. Workflows – No-code assembly using pre-validated nodes

What This Achieves

• Significantly reduces source-code-level attack surface
• Enables rapid development with security gates
• Note: Design-level security (business logic, permissions) still requires human review

Next Steps

• Quick Start - Build your first habit
• Core Concepts - Learn how Habits works
• When to Use Habits - See if it fits your use case