mcat-signup-bot

Bot hosted on EC2 instance to keep checking for MCAT location availabilities. Sends texts and calls to a specified number through Twilio when an appointment is available.

How to use

To use, first create a Twilio account, and get a Twilio phone number as well. Create a secrets.js file that contains the following parameters:

exports.TWILIO_ACCOUNT_SID = ...
exports.TWILIO_AUTH_TOKEN = ...
exports.twilio_number = ...

Create query objects that you want to search (which include an address, date in the format of "Saturday 25th of March 2023", center indices from the search results, and phone numbers to text and call when there is an opening). Create a Bot from bot.js, and pass in the query objects as a list.

Running locally

To run this bot locally, you can simply use node run_bot.js, but this will only run as long as your computer is constantly on.

Setting up an EC2 instance to run this bot (recommended)

You can also run this bot on an EC2 instance, which will allow the bot to continuously run without requiring you to have it on your computer all the time.

Creating EC2 instance

• Create an AWS account, go to the EC2 dashboard, and click "Launch instance".
• Fill in the required fields on the page (make sure you choose free tier!)
  • Name:
  • Application and OS Images: use Ubuntu Server 22.04 LTS
  • Instance type: t2.micro
  • Key pair: create a key pair using RSA with .pem format, and save this to your computer
  • Network settings: default (you can change the SSH traffic to only allow traffic from your IP if you want to be safer)
  • Configure storage: default
• Click 'launch instance'
• After the instance has launched, right-click on the instance and click "Connect"
• Click on the "SSH client" tab of instance connecting, and follow the instructions to SSH into the instance in Terminal
  • On the first time using SSH to this instance, it will ask if you are sure. Type in 'yes' and bypass the prompt.

Transferring bot files to EC2 instance

• Transfer files from local computer using an SFTP software (I use Cyberduck)
  • Fill in server using public DNS (of format ec2-...-amazonaws.com)
  • Username is ubuntu
  • SSH private key is the pem key made earlier
• Transfer the files bot.js, package.json, query.js, secrets.js, and run_bot.js

Installing node and running the bot

• Install node:

$ sudo apt-get update
$ curl -fsSL https://deb.nodesource.com/setup_19.x | sudo -E bash - &&\
  sudo apt-get install -y nodejs

• The terminal may need to restart, and may ask you to select services to restart. Use arrow keys and spacebar to select all, and hit 'enter' to restart the terminal.
• Verify node and npm are installed by running node -v and npm -v, which should print their version numbers
• Install the node packages from package.json:

$ npm install

• Install Chromium, which is needed for Puppeteer to run and create a browser page

$ sudo snap install chromium

• Install the node module "forever", which will allow you to keep the process running after you've logged out of your EC2 instance

$ sudo npm install forever -g

• Run the bot by calling forever start run_bot.js
• Update the forever script and check existing runs by using the following forever commands:

$ forever stopall
$ forever list
$ forever stop run_bot.js
$ forever restart run_bot.js

Notes

MCAT test centers take a while to open up, be patient! It took a while for me to get the ones that I was looking for. I would recommend saving the Twilio number as a contact (and adding it as a favorite) so you won't miss calls, and to always be ready to sign up. Spots don't fill up that quickly (the official MCAT notification system seems to check every hour at most, so most people won't be able to see the openings).

Special thanks to Daniel Shiffman, with a great tutorial on deploying a Twitter bot to an EC2 instance (https://shiffman.net/a2z/bot-ec2/).