Skills are modular capabilities you can use directly with Copilot. Just ask naturally!

Creates automated UI tests for GitHub issues. Creates a HostApp page + NUnit test and iterates until tests actually fail, proving they catch the bug.

Example prompts:

write UI test for issue #12345

create a UI test that verifies Button click updates a Label

Creates XAML unit tests for parsing, compilation, and source generation bugs.

Example prompts:

write XAML test for issue #12345

create test for x:Name resolution bug

Verifies that tests actually catch the bug — tests should FAIL without the fix, PASS with the fix. Auto-detects the test type (UI tests, device tests, unit tests).

Example prompts:

verify my tests fail without the fix on Android

run full verification on iOS

Builds and runs .NET MAUI device tests locally on simulator/emulator.

Example prompts:

run device tests for Controls on Android

run device tests for Core on iOS

Submits and monitors .NET MAUI unit tests on Helix infrastructure.

Example prompts:

run XAML unit tests on Helix

run Core tests on Helix

Builds, packs, and runs .NET MAUI integration tests locally.

Example prompts:

run integration tests

Evaluates test quality, coverage, and appropriateness on PRs that add or modify tests. Also available as a gh-aw workflow triggered by the /evaluate-tests slash command.

Example prompts:

evaluate tests on PR #12345

End-to-end PR reviewer. Orchestrates 3 phases — Pre-Flight, Try-Fix (4 models sequentially), Report. Used by Review-PR.ps1 in the automated pipeline. See PR Agent for full documentation.

Example prompts:

review PR #12345

Standalone deep code review with 345 MAUI-specific review rules distilled from 142 high-discussion PRs. Independence-first — forms opinion from code before reading PR description. Not currently part of the automated PR pipeline — used for manual standalone reviews.

Example prompts:

code review PR #12345

Attempts ONE alternative fix for a bug, tests it empirically, and reports results. Used by pr-review to explore independent approaches across multiple models.

Example prompts:

try an alternative fix for issue #12345

Finalizes a PR for merge — verifies title/description match the implementation and performs a light code review.

Example prompts:

finalize PR #12345 for merge

Finds open PRs that are good candidates for review, prioritized by age, CI status, and review state.

Example prompts:

find PRs to review

Analyzes a completed PR to extract lessons learned — what failed, what worked, what to reinforce. Generates recommendations for instruction files, skills, and documentation. See Learn From PR Agent for the agent that also applies the recommendations.

Example prompts:

what can we learn from PR #12345?

Queries and triages open GitHub issues. Loads a batch and presents issues one at a time for triage decisions with milestone suggestions.

Example prompts:

let's triage issues

grab me 10 issues to triage

Investigates Azure DevOps CI failures — build errors, Helix test logs, and binlog analysis.

Example prompts:

investigate CI failure on PR #12345

Mechanical regression-risk detector that compares removed lines in a PR against additions from recent bug-fix PRs. Labels risks as REVERT, OVERLAP, or CLEAN.

Example prompts:

check PR #12345 for regression risks

Note: Primarily used by the automated review pipeline. Outputs JSON/markdown artifacts.

Auto-labels issues and PRs with area-* and platform/* labels based on technical content, file paths, and platform conventions. Strictly scoped — applies only area and platform labels, nothing else.

Example prompts:

label issue #12345

Note: Also runs as an automated gh-aw workflow.

MAUI-specific dependency flow rules, channel conventions, and feed lookup workflows. Wraps the maestro-cli skill with MAUI-specific guardrails including channel naming rules, safety deny-lists, and input validation.

Example prompts:

feeds for .NET MAUI 10.0

what channels is MAUI on?

promote build to public feed

Runs deterministic .NET MAUI servicing-release and preview readiness surveys. It discovers active release lanes, identifies the commits and source PRs actually shipping with revert detection, classifies regressions and backports, and emits JSON plus Markdown. The skill is report-only and does not cut, merge, tag, or push release refs.

Use the release-readiness-agent instead when you want a synthesized Ready, Conditionally Ready, or Not Ready verdict or a roll-up across active releases.

Example prompts:

release readiness for SR8

how does net11 preview6 look?

give me the raw release-readiness JSON for SR8

Public trackers: Search [Release Readiness] issues