Skip to content

AGENTS.md

Latest commit

 

History

History
39 lines (30 loc) · 1.76 KB

AGENTS.md

File metadata and controls

39 lines (30 loc) · 1.76 KB
Raw

Agent Notes

Guidance for coding agents working in this repository.

Testing samples

The samples/ directory contains many platform-specific projects (Blazor WASM, MAUI, Uno, Avalonia, WPF, WinUI3, MVC, etc.). Most of these require platform SDKs, mobile workloads, or OS-specific tooling (e.g. Windows, iOS/Android workloads, WinUI3) that are not guaranteed to be installed in an agent's local environment. Do not assume dotnet build/dotnet workload for these projects will succeed locally, and do not spend time trying to install missing workloads just to validate a samples change.

The proper way to validate a change to anything under samples/ is:

  1. Push your branch to your local fork (origin), not upstream.

  2. Trigger the Build Samples GitHub Actions workflow via workflow_dispatch against that branch, e.g.:

    git push -u origin <branch>
gh workflow run "Build Samples" --repo <your-fork> --ref <branch>

  3. Poll the run (gh run list / gh run view <run-id>) until it completes, and review the per-job results — in particular the job(s) relevant to the sample(s) you changed, plus the all-samples-built aggregation job.

This builds every sample on its correct runner/OS/toolchain in CI, which is far more reliable than trying to reproduce that locally.

Timing: a full Build Samples run across all sample projects takes approximately 6 minutes end-to-end. Individual jobs (e.g. todoapp-blazor-wasm) typically finish in well under a minute, but slower jobs (Windows, Uno iOS/maccatalyst) can take several minutes each, and they run in parallel. Plan polling/wait times accordingly — don't assume the run is stuck if it's still going after a minute or two; do check back in if it hasn't completed after ~6-8 minutes.