Test Package

SkillDev tools

Run grok test for Datagrok packages and report results

Available today. Use it from your connected AI after setup.

No other account needed.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Test Package skill

What this skill tells your AI

The instructions your AI receives, as published by datagrok-ai/public in .claude/skills/test-package/SKILL.md and read by ahel’s review.

Run grok test for one or more Datagrok packages and provide a summary of results.

Parse the user's input to extract:
- One or more package names (required). These are directory names under public/packages/.
- --host value (default: localhost)
- Use --csv flag to get csv output
- Check if npm install is needed
- Check if package needs to be rebuilt. Compare /dist/*.js timestamp with the most recent timestamp in *.ts or *.js file in the package. If dist folder is newer, use --skip-build
- Any extra grok test flags (--category, --test, --verbose, --skip-build, --gui, --no-retry, --benchmark, --tag)
For each package, run tests from the package directory using the Bash tool:
```
cd /c/Repos/Datagrok/public/packages/<PackageName> && node grok test --host <host> --no-retry [extra flags]
```
Use a timeout of 600000ms (10 minutes). Print output to console.
After all packages finish, provide a summary table with:
- Package name
- Total passed / failed / skipped counts (parse csv output file test-results.csv)
- List of any failed categories and test names
- Overall verdict: PASSED or FAILED
Cleanup csv output file

Advanced