Coffee Break Slackbot

This bot invites randomly picked pairs to a coffee break in Slack each week. As a result, the exchange in the team is increased.

Users that are absent are automatically excluded. Absence is determined by checking if one of the following status emojis is set:

• 🌴
• 🤒

Requirements

• Python 3.9
• jq
• Terraform 1.0.8

Setup

Terraform Backend

Create a Terraform S3 Backend in the AWS account.

Slack App

Create an app in the Slack app directory. Select From an app manifest and choose the workspace for the app. Copy the content of the manifest.yml file into the YAML field and create the app. Then go to the OAuth & Permissions section and install the app to your workspace.

If you like you can give your bot a profile picture.

Configure Lambda Environment Variables

Create a secret named coffee-break-slack-bot/key in the AWS SecretsManager which obtains the Slack app token. You can find the token in the OAuth & Permissions section of your app.

You also need to configure the user list in the SecretsManager. Create a secret named coffee-break-slack-bot/users to store the user list. You can find the id of a slack member in the profile information. Press More and copy the member id. The format of the list has to be ["U1", "U2", ...].

Terraform Variables

If you would like to override default values of variables, copy the terraform.tfvars.dist file to terraform.tfvars and fill in values for the predefined variable names.

List of all variables:

• language: ISO 639-1 language code for the language of the message. Supported are 'en' and 'de'. The default language is english. Unsupported languages fall back to english as well.

Testing

Install requirements before running tests

pip install -r requirements-dev.txt

Deployment

To run terraform apply, first create a session with the AWS Account.

If you use a role run the following command:

aws_credentials=$(aws sts assume-role --profile PROFILE_NAME --role-arn ROLE_ARN --role-session-name "RoleSession" --serial-number MFA_SERIAL_NUMBER --token-code MFA_TOKEN)

If you use your user run the following command:

aws_credentials=$(aws sts get-session-token --profile PROFILE_NAME --serial-number MFA_SERIAL_NUMBER --token-code MFA_TOKEN)

After that export the following variables:

export AWS_ACCESS_KEY_ID=$(echo $aws_credentials|jq '.Credentials.AccessKeyId'|tr -d '"')
export AWS_SECRET_ACCESS_KEY=$(echo $aws_credentials|jq '.Credentials.SecretAccessKey'|tr -d '"')
export AWS_SESSION_TOKEN=$(echo $aws_credentials|jq '.Credentials.SessionToken'|tr -d '"')

Before you run terraform apply make sure that you have installed the dependencies for the lambda function:

bash scripts/build.sh

Navigate to the infrastructure directory, initialize terraform backend and deploy.

cd infrastructure
terraform init
terraform apply

Useful terraform commands

init # initialize terraform backend for selected folder
fmt # format terraform files in this folder
validate # validate the syntax
plan # create plan
apply # create, approve and deploy plan