Slack Auto-Invite Google Cloud Function

Table of Contents

1. Overview
2. Requirements
3. Usage
4. Testing
5. Deployment

Overview

This is a Google Cloud Function that does the following:

• Parses an incoming HTTP POST request from a Marketo contact form Webhook,
• Extracts the real_name and email fields from the incoming request body,
• Automatically invites the user to a given Slack workspace using the Slack Admin User Invite API with the following parameters:
  • The workspace id: SLACK_TEAM_ID
  • The channels to add the user: SLACK_CHANNEL_IDS
• After a successful invitation, sends a chat message to a Slack channel defined by SLACK_MESSAGE_CHANNEL.

Reference Architecture

flowchart TD
    A[User wants to join our Slack Community] -->|opens browser| B(Goes to https://saucelabs.com/community/slack) -->|Fills out Marketo contact form| C(Webhook captures `first_name`, `last_name`, and `email`)
    C -->|HTTP POST | D([fa:fa-server Google Cloud Function])
    D -->|HTTP POST | E[Member added fa:fa-slack, message sent to slack fa:fa-grin]

Loading

Requirements

• A GCP project with the Cloud Functions API enabled.
• A Slack Grid Enterprise account.
• Slack Org/Grid privileges to create an app with the following User Token Scopes:
  • admin.invites.read
  • admin.invites.write
  • admin.users.read
  • admin.users.write,
  • admin.users.invite,
  • chat.write
  • users.read
  • users.read.email
  • users.write
• Node.js and npm installed on your local machine: node v18*

Installation

1. Use nvm to set node to 18 i.e. use nvm18
2. Run npm install
3. Run npm test

Testing

Unit Test are located in ./src/index.test.js, and require the jest testing framework. There are two parts to manually test the functionality:

Test that the Slack App is responding

Below is a Sample Slack API Post Request to the Slack API endpoint

curl -X POST \
-H "Authorization: Bearer $SLACK_USER_TOKEN" \
-H "Content-Type: application/json; charset=utf-8" \
-d '{
    "channel_ids": "$SLACK_CHANNEL_IDS",
    "email": "[email protected]",
    "team_id": "$SLACK_TEAM_ID",
    "email_password_policy_enabled": true,
    "real_name": "Joe Shmo",
    "set_active": false
}' \
https://slack.com/api/admin.users.invite

Test that the Google Cloud Function is able to send the request

Below is HTTP Post Request to the Google Cloud Function URL

curl -m 70 -X POST \
-H "Authorization: bearer $(gcloud auth print-identity-token)" \
-H "Content-Type: application/json" \
-d '{
    "real_name": "Joe Shmo",
    "email": "[email protected]"
}' \
https://<gcp-region>-<project-id>.cloudfunctions.net/slack-auto-invite

Deployment

All GCP deployment resources are located in the terraform directory.

The .gitlab-ci.yml config deploys to GCP and sets up a pipeline that includes several stages:

• test
• build
• deploy

The stages are defined with corresponding jobs, each specifying the container image to be used, scripts to run, rules to apply, and dependencies.

There are several key variables used in the pipeline, including:

• GOOGLE_CREDENTIALS: project service account key
• GOOGLE_PROJECT_ID: google project id
• SLACK_USER_TOKEN: Slack user token with correct API scopes
• SLACK_MESSAGE_CHANNEL: Channel to post notification messages for Slack admins
• SLACK_CHANNEL_IDS: Channels to add the new user
• SLACK_TEAM_ID: The Slack Workspace ID

All of the variables above are crucial to the src/ code and also to validate the terraform configuration.

Deployment Stages

The test stage includes two jobs:

• run-unit-tests: runs unit tests using the Jest framework via npm test.
• run-linter: lints the code located in src/ using eslint.

The build stage has a single job called terraform_validate, which validates the Terraform configuration.

The deploy stage has two jobs:

• terraform_plan: job that creates a deployment plan for Terraform.
• terraform_apply: job that deploys the changes.

The terraform_apply job also has a dependency on the terraform_plan job, and artifacts (the tfplan file) are saved for debugging purporses in case of job failure.