Unclear It is not clear whether the start-of-wait events are also held back by the same flag as the end events.
The background measurement that works out how long a session sits idle now tracks a second kind of state. Alongside periods when work is running, it records when Claude Code is waiting for you to respond, and logs a start and an end for each such wait.
The events marking the end of a run are only sent when an internal flag is on. That flag is off unless it is switched on remotely, and it read off both for this site's account and for an anonymous account.
Separately, Claude Code now logs an event whenever it starts or stops advertising a tool host over its bridge connection.
This is observation only and changes nothing you see. It lets Anthropic see how much idle time is spent waiting on the user, which bears on how long the prompt cache (stored context that makes follow-up requests faster and cheaper) is worth keeping alive.
tengu_dapper_dawn Off in both readingsThe flag server returned off for the account this site reads and for the anonymous baseline. A reading of off cannot rule out a rollout these two readings sit outside of.
This account: off · anonymous baseline: off · compiled default in v2.1.283: off
These values were read against a different version of Claude Code, so treat them as the nearest reading available instead of one taken on this release.
Read once, for one account on one subscription tier, against v2.1.283. It isn't a statement about your account. What a flag value here can and cannot tell you
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.
It is not clear whether the start-of-wait events are also held back by the same flag as the end events.