• A valid GitHub API Token with scope repo
• A valid Slack webhook URL (for development, you can use any webhook testing utility like webhook.site)

This app follows the 12factor convention and uses dotenv to load the configuration.

1. cp .env.sample .env to start with a template file
2. Set the values for GITHUB_API_TOKEN and SLACK_WEBHOOK_URL
3. Repos are configured with env variables starting with REPO_ name, and the content as a serialized JSON for the desired configuration for that repo.

For instance, if we want to see Rails' PRs with label actionpack we need to set up a REPO_RAILS with the following serialized JSON content:

Example:

{
  "name": "rails",
  "org": "rails",
  "includeLabels": [
    "actionpack"
  ],
  "excludeLabels": []
}

You will find this example in the .env.sample, so you can use it as a template to change/setup another repos. Repeat this step for every repo you want to setup.

• name: The repo name in github, the name part in https://github.com/org/name
• org: The org name in github, the org part in https://github.com/org/name
• includeLabels: filter PRs having ALL the labels specified here
• excludeLabels: filter PRs having ANY OF the labels specified here

Emojis are customized via ENV vars starting with EMOJI_. All of them have default values specified in .env.sample. You can use one or many emojis for each section.

• EMOJI_HELLO_MESSAGE
• EMOJI_PR_HEADER
• EMOJI_REPO_HEADER
• EMOJI_APPROVED_REVIEW
• EMOJI_CHANGES_REQUESTED_REVIEW
• EMOJI_COMMENTED_REVIEW
• EMOJI_DISMISSED_REVIEW
• EMOJI_PENDING_REVIEW

• NODE_ENV contains the execution environment, allowed values "development", "production"

QLTY is set up in this project to keep track of maintenance metrics like test coverage and code quality.

This project comes with a group of useful npm run scripts:

• send used to trigger the bot.
• test used to run the tests with Testy.
• coverage used to run the tests and generate a coverage report.
• lint used to run ESLint and report the offenses.
• lint-fix similar to lint but including --fix to autocorrect the offenses.
• open-coverage-report opens the coverage result of the most recent report in a browser.
• generate-dependencies-graph generates a diagram of the modules' dependencies using Madge. This is useful to constantly measuring the complexity of this project.
• churn-vs-complexity gives a summary of the code hotspots (more info in this blog post