Skip to main content
In about 10 minutes you can have an agent that receives Slack messages and replies from your code. Looking for a stack-specific walkthrough? Start with the AI SDK, LangChain, or Mastra quickstart. This page covers the shared onboarding paths (CLI, dashboard, and AI assistant). Prerequisites:
  • A Novu account.
  • A Slack workspace where you can install apps.
  • Node.js on your machine.

Choose your onboarding experience

Pick the path that matches how you like to work. All three get you to the same outcome: an agent connected to Slack that runs your handler code.
Run one command and follow the prompts. The CLI signs you in, creates the agent, connects a channel, and can generate a starter project for you.
You’ll be asked to:
  1. Sign in to Novu (or paste your secret key).
  2. Pick a runtime - AI SDK, LangChain, Custom code, or a managed option like Claude.
  3. Name your agent and connect Slack (or another channel).
  4. Confirm creating a starter project, if prompted.
When you’re done, run npm run dev:novu in the project folder and message your agent on Slack.

Next steps

AI SDK quickstart

Minimal generateText handler for Slack.

LangChain quickstart

Minimal LangChain config for Slack.

Mastra quickstart

Call Mastra generate() from a Novu handler.

Concepts

How agents, handlers, and Novu fit together.

Building blocks

Shared APIs for replies, events, signals, and more.

Agents API reference

Create agents, link providers, and send replies programmatically.