skill-gsb-review

To evaluate model performance after a GSB session, you analyze the model's round responses to determine satisfaction and identify areas needing deeper investigation. This enables precise feedback logging for iterative improvement of model interactions.

The implementation-reviewer agent assesses generated speech behavior by performing detailed professional analysis of post-session model responses. It evaluates interaction quality and decides if adaptive follow-up questions are required for deeper performance investigation.

Yes, you can automatically generate follow-up questions for unsatisfactory model responses. The evaluation mechanism triggers adaptive query generation to produce follow-up prompts, enabling deeper investigation when the initial interaction assessment indicates further questioning is needed.

The best way to log feedback for iterative model improvement is to run a post-session review that integrates an implementation-reviewer agent. This assesses responses, generates adaptive queries if needed, and systematically handles feedback logging for continuous enhancement.

Yes, analyzing GSB model sessions relies on specific path resolver and session ID tool scripts to correctly locate and manage session data. These dependencies ensure the post-interaction evaluation accurately targets the correct model responses for feedback generation.