# Comments on Objects Comments allow team members to leave notes, questions, and feedback directly on objects within InteractiveAI Platform. ### Why Comments Matter Scores capture structured evaluations, but not every observation fits neatly into a numeric or categorical value. Comments fill that gap by providing a space for unstructured feedback. A reviewer might notice that a trace handled an edge case particularly well, or that a prompt version needs refinement for a specific use case. Without comments, this context gets lost or not captured at all. Comments also create a lightweight audit trail. When you revisit a trace weeks later, comments from teammates explain why it was flagged, what was tried, or what conclusions were reached. This institutional knowledge stays attached to the object itself rather than scattered across external tools. *** ### Supported Object Types InteractiveAI supports commenting on the following: **Traces** Add comments to complete execution flows when you want to flag interesting behavior, note edge cases worth investigating, or leave context for teammates reviewing the same trace later.

**Sessions** Leave notes on user interaction patterns across multiple traces. Session-level comments help document recurring behaviors or patterns that emerge over time.

**Prompts** Collaborate on prompt versions by commenting directly on the prompt detail view. Use comments to discuss proposed changes, explain why a version was created, or coordinate with teammates on prompt improvements.