The action.yml file is the metadata file in a custom GitHub Action that defines the main entry point, including information such as the inputs, outputs, description, and the runs key that specifies the main entry point (e.g., a script or a Docker container).

Using a tag is a best practice because tags are immutable and represent a fixed version of your action. By referencing tags, consumers of your action can be assured they are using a stable and specific version of the action, which helps in avoiding issues with breaking changes.

The commit SHA is another reliable way to specify a particular version of an action. By referencing a specific commit SHA, consumers can ensure they are using exactly the code that was written at that moment, avoiding the potential for changes in the future.

GitHub Actions allows you to configure environment protection rules, where you can require specific users or teams to approve the deployment before the workflow proceeds to the next step. This ensures that the required reviewers approve the workflow before any sensitive actions (such as deployment) occur.

Dockerfile : The Dockerfile is required for Docker container actions. It defines the environment for the action, specifying the base image, dependencies, and any commands to set up the action’s runtime inside the container.

action.yml : The action.yml file is required for all GitHub Actions, including Docker container actions. It contains metadata about the action, including the inputs, outputs, and the runtime environment (which in this case is Docker, defined under runs.using ).

Searching the GitHub Marketplace for verified actions published by the cloud provider is the quickest and safest approach. Many cloud providers offer verified GitHub Actions that are maintained and optimized to interact with their services. These actions typically come with the correct configurations and best practices, allowing you to get started quickly without reinventing the wheel.

The GITHUB_ACTOR environment variable indicates the name of the person or app that initiated the workflow. This variable is automatically provided by GitHub in the workflow and can be used to identify the user or application triggering the workflow.

A workflow can be triggered by multiple events either by using array syntax, such as on: [push, pull_request], or by configuring each event separately when filters or activity types are needed. Option C is valid because both push and pull_request are supported workflow events. Option E is also valid because it configures two event triggers: push filtered to the main branch and release filtered to the created activity type. Option B is invalid because commit is not a GitHub Actions workflow event. Option A wrongly puts environments under on. Option D defines branch filters without an event. Option F uses initiate, which is not a valid event. GitHub documents both multiple-event array syntax and separate event configuration for filters.

Since the self-hosted runner is configured on a Linux VM and the script is written in PowerShell, you can invoke the script using the pwsh (PowerShell Core) shell in a run step in the workflow. This ensures that the script runs as-is on the Linux runner, as PowerShell Core (pwsh) is cross-platform and supports Linux.

The workflow_dispatch event is used to manually trigger a workflow run in GitHub Actions. You can specify this event in the workflow file to allow users to manually trigger the workflow from the GitHub UI, often with optional input parameters.

The GET /repos/:owner/:repo/actions/runs/:run_id/logs API endpoint is used to retrieve the logs of a specific workflow run identified by run_id. This is the correct method for downloading logs from a workflow run.