– Other –
It would be great if hook inputs could include token usage details (prompt tokens, context tokens, completion tokens, total).
This would enable:
Cost tracking
Efficiency analysis across prompts, context, and agents
Even basic token breakdowns would be very helpful. Thanks!
Hey, thanks for the request. This is a great idea for improving hooks.
Right now, hook inputs don’t include token metadata, but I can see how that would be useful for cost tracking and efficiency analysis. The team is already working on hook improvements.
I’ll pass your feedback to the team. A similar request came up about adding agent mode to hook input data:
Thanks for the update, appreciate you passing this along.
For power users who spend more and more per month, even small token-efficiency gains start to matter a lot. Having basic token metadata in hook inputs would make cost tracking and optimization much easier.
Looking forward to the upcoming hook improvements.