MTG Slack Card Fetcher

A card fetcher that utilizes Slack's Realtime Messaging API

This bot provides a functionality that you might have seen before on sites like Reddit and MTGSalvation. The functionality is straight forward: wrap a card's name in double brackets and the bot will respond with the cards (e.g. [[Black Lotus]]). See the pictures below for an example.

For the application code, this project uses Node.js, MongoDB and the slackbots npm package. For deploying the bot, the project uses Packer to build an AMI with the dependencies and Terraform to start the instance on AWS. The data for the cards from the mtgjson GitHub repo.

Requirements

• AWS Account, API key, and secret
• MongoDB
• Terraform
• Packer
• Node.js
• A Linux machine of some sort

Setup

For each of the following setups, you must perform this setup:

1. Setup your integration on Slack.
2. Goto https://[YOUR_TEAM].slack.com/apps/manage/custom-integrations.
3. Click Bots.
4. Click Add Configuration.
5. Setup your bot.
6. Get your API Key.
7. Clone the MTG JSON repository somewhere else on your computer.
8. Create a config.js in the secrets directory. Use the Secrets Example below.
9. npm install the dependencies for this project.
10. Run the move-cards scripts to move all the English file from the MTG JSON json directory to the data directory in this repo like so: PATH_TO_JSON=/path/to/mtgjson/json PATH_TO_DESTINATION=/path/to/data/directory node ./tasks/move-cards.js

Local

1. Start MongoDB
2. Add your mongo url to the mongoUrl field in the config/secrets.js
3. Add the collection name to the collectionName field in the config/secrets.js
4. Fill in the bot fields in the config/secrets.js. The bot ID may be found in the URL when you're editing the bot on Slack's website (something like https://planeswalker.slack.com/services/BOT_ID_HERE)
5. Ingest the card data with the following scripts: PATH_TO_JSON=/path/to/data/directory ./tasks/ingest_cards
6. Invite the bot to the Slack channel you want it to listen to.
7. Fill in the adminUser field in the secrets Run the index script: node ./index.js
8. You'll get a PM from the bot saying it is fired up.

AWS Deployment

This repository also comes with Terraform and Packer resources for easy AWS cloud deployment. These directions will teach you how to run your very own instance of the bot in the cloud with just a few commands.

In my personal repository, I store secrets and other private information inside of the Terraform and Packer config files, so you'll have to follow the steps and examples below to set up your own configurations.

Setup

1. Make sure that your AWS_SECRET_ACCESS_KEY and AWS_ACCESS_KEY_ID are in your environment.
2. Install Packer and Terraform. If you have brew, run brew install terraform and brew install packer.
3. Create a directory outside of the project for your state file. The path to this directory will be used later on during the deploy.
4. Create a SSH keypair and add it to your SSH agent. For instructions on this, check out GitHub's helpful walkthrough.
5. Create a deploy/config/keys directory and move the SSH keypair into it.
6. Inside of the deploy/config directory, create a .tfvars file. Make sure this file's name corresponds with the name of the whatevever you name your state file. State files are created during the terraform apply process.
7. Copy the Terraform Variables Example below into the .tfvars file and populate each of the variables with the correct values. Use the example values if you're not sure what to put.
8. Create a .json file inside of the deploy/config directory to store the AMI configuration. Name the file such that it matches the name of your .tfvars file except with an ami- on front (e.g. ami-us-east-2-dev.json if you've name your .tfvars file us-east-2-dev.tfvars)
9. Copy the Packer Config Example below into the .json file to, and populate each of the variables with the correct values. Use the example values if you're not sure what to put. Note: the source_ami will vary by region.

Below is the recommended structure for your deploy/config directory:

config
  ├── ami-us-east-2-dev.json
  ├── keys
  │   ├── aws_mtg
  │   └── aws_mtg.pub
  └── us-east-2-dev.tfvars

Deploy

Now that you have the configurations in place, deployment is very easy. Run the below commands to build the AMI and apply the terraform resources (except replace the NAME_OF_VAR_FILE with the actual name of your .tfvars file):

cd deploy && MTG_STATE=/path/to/state/file/directory ./utils/pack NAME_OF_VAR_FILE && ./utils/build apply NAME_OF_VAR_FILE

For example, if the name of your .tfvars is us-east-2-dev. Running the below command verbatim will work:

cd deploy && MTG_STATE=/path/to/state/file/directory ./utils/pack us-east-2-dev && ./utils/build apply us-east-2-dev

When it's done packaging the dist, building the AMI, and applying the terraform resources, you'll get a PM from the bot saying it's fired up and ready to go.

Secrets Example

module.exports = {
  token: 'SLACK API KEY',
  mongoUrl: 'MONGO URL',
  collectionName: 'COLLECTION NAME',
  adminUser: 'YOUR SLACK USERNAME',
  bot: {
    id: 'BOT ID',
    icon: 'BOT EMOJI ICON',
    name: 'BOT NAME'
  }
};

Terraform Variables Example

aws_region = "AWS REGION (e.g. us-east-2)"
slackbot_instance_type = "EC2 INSTANCE TYPE (e.g. t2.micro)"
environment = "NAME OF ENVIRONMENT (e.g. dev)"
public_key_path = "RELATIVE PATH TO PUBLIC KEY FOR KEY PAIR (e.g. config/keys/aws_mtg.pub)"
private_key_path = "RELATIVE PATH TO PRIVATE KEY FOR KEY PAIR (e.g. config/keys/aws_mtg)"
key_name = "KEY NAME (e.g. mtg_slackbot)"
white_listed_ip = "YOUR PUBLIC IP"

Packer Config Example

{
  "node_version": "7.6.0",
  "region": "REGION TO BUILD AMI (e.g. us-east-2)",
  "source_ami": "AMI TO BUILD OFF OF (e.g. ami-fcc19b99)",
  "instance_type": "INSTANCE SIZE (e.g. t2.micro)",
  "environment": "ENVIRONMENT NAME (e.g. dev)"
}