All what we need to do is to add a YAML file in the .github/workflows folder of the project with the instructions.

In this case I added the following file:

examples/test2-harness-ui-ci.yml

name: Perl

on:
  push:
  pull_request:

jobs:
  test:
    strategy:
      fail-fast: false
      matrix:
        runner: [ubuntu-latest] # , macos-latest, windows-latest
        perl: [ '5.32' ]

    runs-on: ${{matrix.runner}}
    name: OS ${{matrix.runner}} Perl ${{matrix.perl}}

    steps:
    - uses: actions/checkout@v2

    - name: Set up perl
      uses: shogo82148/actions-setup-perl@v1
      with:
          perl-version: ${{ matrix.perl }}
          distribution: ${{ ( startsWith( matrix.runner, 'windows-' ) && 'strawberry' ) || 'default' }}

    - name: Show Perl Version
      run: |
        perl -v

    - name: Install Modules
      run: |
        cpanm -v
        cpanm --installdeps --notest .

    - name: Run tests
      run: |
        perl Makefile.PL
        make
        make test

A quick explanation:

The name of the file

I used in the repository was ci.yml, but any name would do and I think both the yml and yaml extensions work.

Name

name: Perl

There must be a name field, but the content is anything you like.

Trigger

on:
  push:
  pull_request:

Means the job will run on every push and on every pull-request.

Jobs

There can be a number of jobs, the names of the jobs can be anything. In our case we have one job and I called it test

jobs:
  test:

Strategy and Matrix

    strategy:
      fail-fast: false
      matrix:
        runner: [ubuntu-latest] # , macos-latest, windows-latest
        perl: [ '5.32' ]

The matrix keyword allows us to define several variables (We used the words runner and perl in this case) to define a list of values. This allows us to run the tests in various conditions. In this case I configured the tests to run on Ubuntu only and on perl version 5.32, but later both can be extended.

The fail-fast parameter tells Github Actions what to do if one of the jobs in the matrix fails. false means to run all the jobs to the end. true would mean to cancel all other jobs if one of the jobs derived from the matrix fails.

The platform

The runs-on keyword is using one of the variables from the matrix to select on which platform to run on. The name uses both variables from the matrix, but it is only used for logging.

    runs-on: ${{matrix.runner}}
    name: OS ${{matrix.runner}} Perl ${{matrix.perl}}

steps

In each job there can be a number of steps.

step: Check out the current repository

The first step is to check out (clone) the current repository.

    - uses: actions/checkout@v2

step: Set up Perl

    - name: Set up perl
      uses: shogo82148/actions-setup-perl@v1
      with:
          perl-version: ${{ matrix.perl }}
          distribution: ${{ ( startsWith( matrix.runner, 'windows-' ) && 'strawberry' ) || 'default' }}

Here we use to pick the version of perl using the variable from the matrix.

step: Show Perl Version

    - name: Show Perl Version
      run: |
        perl -v

We display the version of the actually installed perl. Just so we will be able to verify it, if needed.

step: Install dependencies

We use cpanm to install the dependencies. We use the --notest flag to reduce the time spent on installing the dependencies.

    - name: Install Modules
      run: |
        cpanm -v
        cpanm --installdeps --notest .

step: Run tests

Finally we have the step to run the tests:

    - name: Run tests
      run: |
        perl Makefile.PL
        make
        make test

What to do now?

Once the file is added to the repository (once my pull-request is accepted) on every push GitHub will run this workflow and it will send the author an email if anything fails with a link to the logs.

There are, of course many other options one can use, but as a first step it is great to get feedback on every push.

The workflow will also run on every pull-request, meaning that the author of the pull-request will be able to get feedback even before the author has time to review the changes.