Create a Slackbot in under 10 minutes using Node.js – Sinch MessageMedia

Chatbots — they’re cool, fun and useful virtual assistants to have. In this tutorial, we’ll build a basic ChatBot that can reply to queries via Slack. We’ll be running this on our local machine and to make it even simpler, we’ll be leveraging an awesome tool called Botkit which will do most of the heavy-lifting for us.

Prerequisites

Node
Slack App
10 Minutes

Setting Up

Create an empty folder for your project, let's name it testApp.

Run the code – npm init -y
This command quickly generates a package.json file.

Once that’s created, run the following commands – npm install dotenv
and npm install botkit
The first package adds support for environmental variables in your Node project and the second package adds the botkit tool to your project.

Adding Bots to Slack app

Sign in to your workspace
Once signed in, you need to go the App manager screen. This can be found at https://xyz.slack.com/apps/manage (replace xyz with the name of your workspace)

Now click on App Directory at the very top and in the search bar that appears, search for “Bots”
Select the first one that says “Connect a bot to the Slack Real Time Messaging API”

Click on the “Add configuration” button

Once you’ve named your bot, you should be redirected to a page that displays an API Token

Make note of this API token as we’ll be using this in the next section

Coding the SlackBot

Create a .env file inside the “testApp” folder created earlier, open it up in your editor and add your API token in it. My file looks like this:

Create an index.js file inside the same folder and open it up in your editor. Write down the following code. I've commented it so you can understand what's going on with each line:

// Add support for environmental variables
require('dotenv/config');

// Import botkit package
var Botkit = require('botkit');

// Create a controller for the Slackbot
var controller = Botkit.slackbot({
  debug: false
});

// Use the token to connect the bot to your Slack app
controller.spawn({
  token: process.env.TOKEN
}).startRTM();

// Determine behaviour of the Slackbot
controller.hears(['hello', 'hi'], ['ambient', 'direct_mention', 'mention'], function (bot, message) {
  bot.reply(message, "Hello.");
});

Let’s try and decode the last bit of code. FYI, this is where all the magic happens. We’ll breakdown the function hears into 3 parts:

Trigger: ['hello', 'hi']
Context: ['ambient', 'direct_mention', 'mention']
Callback – What is the bot supposed to do?

The triggers aren’t exact matches, so if the combination of letters occurs, the bot will fire the function. You can use regex if you’d like exact matches.

The context basically means where the bot heard the message. Was the bot mentioned or did someone send the message in a channel? The list of options to choose from are:

ambient: message came from a channel
direct_mention: the bot was mentioned in the message (@bot)
mention: the bot was mentioned by name
direct_message: message received directly via private chat

Are you excited to see your Slackbot in action? Let’s find out together.

Lights, camera, Slackbot

If everything went well, you should see it online in the sidebar under Apps.

Bots are shy, so you need to invite them your channel. Try inviting your bot to the general channel with the /invite command in Slack. In my case I had to write: /invite @robobot

Now fire up your command line and run the node app.

Let’s send a “Hi” message in the general channel.

Conclusion

Well done – you’ve successfully built a Slackbot that’s capable of replying to certain trigger words. Now you can add more complex logic to the bot and get it to do much more than just saying Hello. The possibilities are endless. Thank you for reading!

Get the code

All the code for this tutorial is available in the Slack Node.js Tutorial Github repo

Resources

Botkit Github repo