# Yesterday I created a build and push artifact CI/CD pipeline for AZ-400 practise in Azure DevOps!

This weekend I spent some time improving my small .NET helper library CI/CD pipeline as part of my AZ-400 study path. The goal wasn't just to pass the exam, but to build something I can reuse and point to in real projects.

**What I set up:**

- A small .NET extension-library.
- AI assisted unit test generation with Aider and Qwen2.5
- An Azure DevOps YAML pipeline that builds from a GitHub repo.
- Automated NuGet packaging and publishing to GitHub Packages using a GitHub PAT.
- NuGet restore caching to speed up builds.
- Static check style gate with `dotnet format --verify-no-changes`.
- Code coverage
- Automatic versioning using `$(Build.BuildId)` so every CI run produces a unique SemVer-compatible package.

**A few small but meaningful lessons:**

- Adding `paths.exclude: - 'azure-pipelines.yml'` to the trigger prevented the pipeline from running on every YAML edit and gave me intentional control over when to run it.
- Aider helped scaffold most unit tests, but I ended up tweaking and fixing them manually, proving that AI-assisted development is best when paired with human review. The pipeline now reports ~63% line coverage, which is enough to show the tooling works: the point is the pipeline, not the number.
- I also hit a caching quirk: the Cache@2 post-job task kept failing because the target cache directory did not exist on first runs. A quick `mkdir -p ~/.nuget/packages` before the cache task fixed it and left the pipeline green.

**Next steps:**

- Recreate the pipeline in GitHub Actions
- Add SonarQube

![Azure DevOps pipeline run screenshot](https://images.koorevaar.com/linkedin-import/az400-azure-devops-pipeline.jpg)

---

*Co-authored with Claude.*

If this was useful, [you can buy me a coffee](https://ko-fi.com/p47k0).

New articles: [follow via RSS](https://blog.koorevaar.com/feed.xml).