Pushing small GitHub Actions changes just to see whether a workflow works is slow, noisy, and burns hosted runner minutes unnecessarily ❌
Act solves a lot of that pain by running your GitHub Action workflows locally. It shortens the feedback loop when you are iterating on workflow logic, shell steps, environment variables, or basic job wiring.
❌ Figure: Bad example - Using hosted runners as your first feedback loop (lots of tokens burned)
✅ Figure: Good example - Use act to tighten the workflow feedback loop. Bottom 5 lines show success (no tokens burned)
Use Act in a pre-push hook for quick CI checks such as builds and tests, but avoid running deployment or other side-effecting actions locally.
Act gets you closer to your final workflow much earlier, but it is still not identical to running on GitHub's hosted runners as it emulates a GitHub Actions like environment. Some workflows behave differently depending on the runner image, available tools, service containers, or platform-specific behaviour.
That means Act should improve your inner loop, not replace your real CI run. You can read more about Act via the Act documentation.