Nothing to try yetTier: how much it should matter to you
2Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
Output LimitsArea: what it touches
In DevelopmentKind: in v2.1.305,
In DevelopmentSection of the release
What
When the model's reply runs into its maximum length, Claude Code nudges it to carry on and retries. It used to choose between two fixed nudges. A server switch can now set, separately for the main conversation and for subagents (helper copies of Claude):
which nudge to send, with a new option for when there was no visible output
how many retries to allow before stopping early
When the server sends nothing, behaviour stays as before. Stopping early is now recorded as its own outcome.
Why
Nothing changes until the server sends a value. It gives a way to tune how hard Claude Code pushes to recover from a cut-off reply.
Read from
Feature flag
tengu_tingly_pikeNot enough to say
Nothing here resolved what this flag was doing on this version, so nothing here should be read as on or off.
This account: no value returned · anonymous baseline: no value returned · compiled default in v2.1.305: not a boolean we can read
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.
Flag reading movedThe flag server now returns null for tengu_tingly_pike, read as this 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.
Flag reading movedThe flag server now returns nothing at all, having stopped carrying it for tengu_tingly_pike, read as this 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.
