Before Claude Code runs a shell command, it checks whether the command is safe to run without asking you. One thing it looks for is text of the form NAME=value, which can set an environment variable. Environment variables are named settings that programs read. Before, any NAME=value whose name was not on a known-safe list made the check fail.
Now a NAME=value that comes after the command word of certain allowed commands is no longer treated as an unsafe variable setting, as long as the name fits a set pattern. This sits behind a remote switch, and the code falls back to on when the server sends no value.
Commands such as cmd KEY=val should lead to fewer permission prompts or fallbacks.
tengu_quiet_larkspur Not enough to sayNothing 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 v2.1.285: on
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 v2.1.285. It isn't a statement about your account. What a flag value here can and cannot tell you
Which commands this covers and which names the pattern accepts are not known.
New in this build: tengu_quiet_larkspur