Claude Code can pull useful facts out of a conversation and save them as memory for later sessions. When one of these saving passes ends, Claude Code now reports how it went, as one of these outcomes:
- saved
- unfinished
- writes refused
- nothing to save
- failed
The outcome is also reported when saving runs through MCP (a standard way for Claude Code to connect to outside tools) and when the pass hits an error. The report goes to an optional listener, so it only has an effect when a program built on Claude Code chooses to listen for it.
Programs that embed Claude Code can now see whether memory saving actually worked, without having to guess.
tengu_bramble_lintel 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 v2.1.295: 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.
Read once, for one account on one subscription tier, against v2.1.295. It isn't a statement about your account. What a flag value here can and cannot tell you