agent-observability-experiment-analyzer

Analyzes individual or paired LLM experiments to produce evidence-backed findings and recommendations.

When to use it

Use when asked to analyze an experiment, compare experiments, or evaluate against a baseline using one or two experiment IDs.

Give it one or two experiment IDs; it asks about analysis scope and delivery, then produces a report in chat by default.