What it does
The active rule path asks Noul yes-or-no questions. Code supplies the finding text, locations and failure policy; an unflagged change is not proof of correctness.
Agent tooling
LintPal
Write review rules in Markdown and ask Jev whether changed lines violate them; code attaches findings to those lines.
Bookmark: LintPal
Keep this in your collection.
Leave a noteWhat would you try with this?
: LintPal
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Start with one written rule and a small code change before adding automatic checks. A developer can turn your team’s requirements into questions and review the flagged lines. The rules and relevant code go to an outside AI service. A change that receives no warning is not proven correct.