I ve been a technical educator for a long time, but this is the first agent skill I m sharing publicly.Run it and it takes the code you re about to push (or point it at a commit), prioritizes what to review, then probes your reasoning, assumptions, understanding of consequences, etc.At the end it generates a small report you can paste into a PR description for your reviewer or keep as a personal record.The goal is to push code only once you can be accountable for it. Useful to maintain your own understanding or have your junior engineers use.I ve enjoyed using it personally.