Feature request: Inspectable context and activity view for ChatGPT agent chats

I’d like an inspectable view for each ChatGPT agent chat or run that shows:

  • which user-configured instructions applied;
  • which tools and connected apps were available and which were used;
  • which files and sources were accessed; and
  • what actions were taken, with timestamps and results where available.

This would help users understand, verify, and troubleshoot agent behavior. I’m requesting an operational record of the system’s context and actions, not private internal reasoning.

1 Like