Version updated for https://github.com/yashchimata/judgegate to version v0.1.0. This action is used across all versions by 0 repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. Action Summary judgegate is a CI trust gate for LLM judges that measures their reliability against human labels using Cohen’s kappa and bootstrap confidence intervals. It helps teams avoid relying solely on AI judgments by verifying their accuracy and identifying noisy or unreliable judges. The action verifies a judge, asks the label budget question, stops labeling early, and validates any labels file before running.