mkdev-me/terraform-aws-github-runner-lambda-microvms

When you use GitHub to run automated tests or deployment tasks (called GitHub Actions), those tasks run on machines provided by GitHub. GitHub charges for that compute time and limits how much you can use on free plans. This Terraform module lets you run those same tasks on your own AWS account instead, using a service called Lambda MicroVMs. A MicroVM is a lightweight, isolated mini-computer that starts up very quickly from a saved snapshot. When a GitHub task is queued, a webhook triggers one of these mini-computers to spin up, run the task, and shut itself down completely. Because each task gets a fresh, isolated machine that disappears when done, you never pay for idle time and each run is clean. Terraform is a tool that lets you describe your cloud infrastructure as configuration files. With this module, you write a few lines of configuration and run one command, and it builds the MicroVM image, sets up the webhook receiver, and wires everything together in your AWS account. GitHub webhooks can be configured automatically as well. There are a few requirements before using it. You need an AWS account, the Terraform command-line tool, and a GitHub App, which is a type of machine credential you create in GitHub's developer settings. The MicroVMs only run on ARM-based processors (AWS Graviton), so jobs must specify that. The first setup takes a few minutes because AWS has to build the machine image. Docker-based jobs are supported. The main benefit is cost. Ephemeral, single-use runners that only exist while a job is running are generally cheaper than GitHub's hosted runners for teams with high CI usage. There is no idle machine to manage or pay for between jobs. This module is aimed at engineers and teams who already use Terraform to manage AWS infrastructure and want to cut GitHub Actions costs with minimal ongoing maintenance.