[Suggestion] Expose usage-limit reset time to the agent so it can self-schedule retries

When a session hits the plan's usage limit, tool calls fail with a message like "[model] is temporarily unavailable, so auto mode cannot determine the safety of [tool] right now. Wait briefly and then try again." This carries no structured/machine-readable reset time — just a vague "wait briefly."

In practice, the reset appears to follow a fixed-duration window (empirically observed as ~5h) rather than a rolling cooldown from the moment the limit was hit — but this has to be inferred by watching for a boundary and remembering it, rather than being told.

Request: when a tool call is blocked for this reason, include a structured reset timestamp (or remaining duration) in the response. That would let the agent compute an accurate retry time and self-schedule a resume (e.g. via existing cron/wakeup tooling) instead of blind-retrying, giving up, or relying on user-supplied, empirically-inferred cadence math. It would also let the agent tell the user precisely "blocked until X" instead of "try again later."

Imagined by myself, drafted by Claude AI.