AI agents in widgets show the full reasoning and all tool call details inline, making conversations hard to follow. The current design of Port’s AI chat surfaces tool calls and execution details (now grouped into collapsible sections) to keep transparency while reducing clutter, but there is no user-facing control in the widget to hide or minimize this behavior further.

The request is to add configuration options (at least at widget level, ideally per user preference) to:

Hide or significantly minimize the “thought process” and tool call logs, showing primarily the final answer.

Choose whether tool call sections are shown, collapsed, or completely suppressed in the main conversation view.

This would let teams keep the current transparent, tool-aware experience where needed, while offering a “clean mode” for end users who just want concise answers without detailed execution traces.