Copilot Review Loop

Purpose

Automate an iterative Copilot-driven review loop: trigger a GitHub Copilot review, address its feedback one comment at a time, then re-trigger Copilot to review again. Repeat up to 2 cycles until all critical issues are resolved.

This is an in-session loop — one long agent context that iterates internally. It is not the Ralph Wiggum pattern, which is an external shell loop that spawns a fresh agent session per comment. For that, see the ralph-wiggum-loop skill instead.

Typical invocations

Users trigger this skill with prompts like: