How to use GitHub Copilot for boilerplate and setup

Creating new projects is rarely the most exciting part of development work. Environments need to be set up, repetitive configuration must be written, and standard structures have to be created before the actual logic can begin.

This is exactly where an AI tool like GitHub Copilot can make your life easier—not by writing algorithms, but by removing friction in the setup process.

Example 1: a new Flask app in minutes

Create an empty folder in Visual Studio Code and open the Copilot Chat panel (or use the command line in the editor’s built-in terminal). Then type:

copilot: create a new Python Flask API with one GET route /status returning JSON {"status": "ok"}

1. Copilot typically generates:

• app.py with a minimal Flask structure
• an appropriate .gitignore
• a suggested requirements.txt

2. Next, use Edits mode in the Copilot panel with the prompt:

“Add a Dockerfile for this project using Python 3.12 and expose port 8080.”

Copilot will usually generate a usable Dockerfile with pip install -r requirements.txt, EXPOSE 8080, and CMD ["python", "app.py"].

Copilot does not always guarantee fully working code; this depends on the prompt and the model. That’s why it is always important to review and test the generated code.

3. Then add configuration:

“Generate a docker-compose.yml that maps port 8080 and mounts the current directory.”

Copilot usually understands this and produces a working setup, ready to run with docker compose up, although small adjustments may be necessary.

In just a few minutes, you’ve gone from an empty folder to a running prototype—without digging through old snippets or documentation.

Example 2: Node.js REST API with Express

The same approach works across languages. In a new Node project, you can for example write:

copilot: initialize a new Express.js project with a single GET endpoint /api/ping returning JSON {"pong": true}

Copilot now automatically creates:

• package.json with the necessary dependencies (express, nodemon)
• server.js with a basic Express setup
• a .gitignore that includes node_modules

In Edits mode, you can then extend the setup:

“Add a Dockerfile for Node 20 using port 3000 and include a start script.”

Copilot generates a complete Dockerfile with WORKDIR /usr/src/app, COPY package*.json ., RUN npm install, EXPOSE 3000, and CMD ["npm","start"].

Finally, add:

“Generate a simple README.md describing how to run the container locally.”

You now have a complete, containerized Express project in under five minutes - with documentation, dependencies, and a build file included.

Example 3: automate the build process with CI scripts

Copilot can also save time on small, repetitive DevOps tasks. Instead of writing build scripts from scratch, you can use Edits mode to generate a standardized CI workflow.

You can, for example, prompt Copilot with:

“Create a GitHub Actions YAML workflow for building and testing a Python Flask app using pytest.”

Copilot will typically suggest a workflow file such as .github/workflows/build.yml with steps like:

• checking out the code (actions/checkout@v3)
• setting up Python (actions/setup-python@v4)
• running tests (pytest)

This sets up the environment and runs tests automatically. In practice, Copilot will often generate a full workflow, but the key value is that you quickly get a working starting point that you can refine.

With a few adjustments, you have a CI setup that runs tests on every commit—without writing the entire YAML file manually.