Version updated for https://github.com/cholmess/breakpoint-ai to version v0.2. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. Action Summary BreakPoint AI is a GitHub Action designed to automate the detection and prevention of risky changes to large language model (LLM) outputs before deployment. It evaluates modifications to prompts or models by comparing new outputs against a stored baseline, identifying issues like cost increases, PII leaks, or output format regressions that traditional unit tests might miss.

Version updated for https://github.com/cholmess/breakpoint-ai to version v0.1.8. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. Action Summary BreakPoint AI is a tool designed to evaluate and prevent issues in AI model outputs before deployment, ensuring cost efficiency, format integrity, and security compliance (e.g., detecting Personally Identifiable Information). It automates comparisons between baseline and candidate outputs, providing deterministic evaluations with clear statuses (ALLOW, WARN, or BLOCK) and actionable feedback.

Version updated for https://github.com/cholmess/breakpoint-ai to version v1. This action is used across all versions by ? repositories. Action Type This is a Composite action. Go to the GitHub Marketplace to find the latest changes. Action Summary BreakPoint AI is a GitHub Action designed to evaluate AI model changes and prevent problematic deployments. It automates the detection of issues such as increased costs, sensitive data leakage, and output regressions by comparing a baseline and candidate model, providing a clear status (ALLOW, WARN, or BLOCK).