Skip to content

10.CopilotCLIIntegration.md

Latest commit

 

History

History
93 lines (64 loc) · 3.24 KB

10.CopilotCLIIntegration.md

File metadata and controls

93 lines (64 loc) · 3.24 KB
Raw

GitHub Copilot Lab

Why Integrate Copilot CLI?

Copilot CLI supports authentication with personal access tokens (PATs), which makes it easy to run inside CI/CD environments such as GitHub, GitLab, Jenkins, and Azure DevOps. Integrating the CLI lets you bring Copilot’s assistance into these pipelines and extend how you use Copilot beyond the editor.

How We'll Use It in This Lab

In this lab we will:

  • Integrate Copilot CLI into a GitHub Actions workflow so every new PR triggers an automated review.
  • Use Azure OpenAI in the same workflow to summarize the Copilot CLI findings and post them back to the PR.
  • Note: the example focuses on code review, but you can customize the workflow for other scenarios.

Lab Environment

Software Requirements

  • Node.js: >= 22.0.0
  • npm: >= 10.0.0
  • VS Code: latest version
  • GitHub Copilot: signed in

Lab Steps

Step 1: Create a PAT

1.1 Goal

Generate a personal access token.

1.2 Procedure

  1. Sign in to GitHub Go to Settings. launch copilot

  2. Create the PAT Navigate to Developer settings → Personal access tokens → Tokens (classic), click Generate new token, and store it securely. launch copilot

1.3 Validation

  • The token is generated and saved locally.

Step 2: Create GitHub Secrets

2.1 Goal

Add secrets to the repository.

2.2 Procedure

  1. Open the repository settings On GitHub, open the target repository and go to Settings.

  2. Create secrets Under Secrets and variables → Actions, click New repository secret and add the following (use the PAT from Step 1 for COPILOT_CLI_PAT):

    COPILOT_CLI_PAT
AZURE_OPENAI_API_KEY
AZURE_OPENAI_BASE_URL
AZURE_OPENAI_MODEL

    configure secret

2.3 Validation

  • The secrets are created successfully.

Step 3: Integrate Copilot CLI and Azure OpenAI in GitHub Actions

3.1 Goal

Wire Copilot CLI and Azure OpenAI into the GitHub workflow.

3.2 Procedure

  1. Open the local repo in VS Code Make sure you are still on the feature/testbranch branch.

  2. Add the GitHub Actions workflow Create .github/workflows/pr-copilot-review.yml. Use the example at https://github.com/jacwu/github-materials/blob/main/.github/workflows/pr-copilot-review.yml as a reference.

  3. Add the Azure OpenAI helper script Create scripts/normalize_review_result.py. Use https://github.com/jacwu/github-materials/blob/main/scripts/normalize_review_result.py as a reference.

  4. Commit and push Commit your changes locally and push them. This activates the earlier PR from Step 2 and triggers the new GitHub Actions workflow, which runs the Copilot review and posts the summarized results.

3.3 Validation

  • The Actions tab shows the workflow run. trigger action

  • Within the run details, both Run Copilot Code Review and Normalize Review Table with Azure OpenAI succeed. action running

  • The PR displays the review summary. Because this commit adds only YAML and Python scripts, the review might not flag issues; push problematic code to experiment further. review result