Version updated for https://github.com/hidai25/eval-view to version v0.7.0. 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 EvalView is an open-source GitHub Action designed to detect and manage silent regressions in the behavior of AI agents, such as changes in tool usage, output quality, or decision-making processes. It automates tasks like tracking behavior drift, classifying changes, and facilitating safe auto-healing of issues, providing graded confidence levels instead of binary alarms to distinguish between provider/model updates and actual regressions. This tool empowers developers and teams to ensure their AI agents continue to function correctly and reliably without requiring extensive resources.

Version updated for https://github.com/hidai25/eval-view to version v0.6.2. 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 EvalView is an open-source GitHub Action designed to detect silent regressions in AI agents by monitoring and analyzing changes in their behavior, tool usage, and outputs. It automates tasks such as behavior snapshotting, drift detection, change classification, and safe auto-healing, enabling developers to identify and address issues caused by model updates or system changes before they impact users. Its key capabilities include tracking behavior consistency, distinguishing between provider/model drift and system regressions, and streamlining the evaluation-review-deployment cycle for AI agents.