The system prompt is the set of instructions Claude Code sends to the model at the start of a conversation. Two of its sections can be switched on or off remotely by Anthropic:
- a section about reduced delegation, which is on unless it is switched off remotely
- a section about tool parameters, which is off unless it is switched on remotely
Both switches existed before. The code that reads them was reorganised, but apart from internal renaming there is no clear difference in how these sections are included.
In practice, nothing about these two sections appears to have changed, and there is nothing you need to do.
tengu_slate_bittern 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 v2.1.310: 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.
tengu_silent_harbor 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 v2.1.310: 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 v2.1.310. It isn't a statement about your account. What a flag value here can and cannot tell you