If you're running integration tests for GitHub Copilot for Azure skills and need to file bugs when they fail, this automates the tedious part. Point it at your JUnit XML output and it'll parse the failures, dig through your test run logs to find the agent metadata, pull the actual failing line of code from your test files, and create a properly formatted GitHub issue in the microsoft/github-copilot-for-azure repo. The standout bit is that it requires a diagnosis before filing, forcing you (or Claude) to actually compare what the agent did versus what the test expected and suggest whether the fix belongs in the skill, the test, or the fixtures. It's opinionated about including full untruncated logs in collapsible sections, which is the right call for debugging later.
npx -y skills add microsoft/github-copilot-for-azure --skill file-test-bug --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
wshobson/agents
dbt-labs/dbt-agent-skills
github/awesome-copilot