Claude Code reportedly shows summarized thinking as assistant messages
A user says Opus won't recognize what appears to be its latest message when asked if it said it.
TLDR
A user claims Claude Code displays summarized thinking as assistant messages. They say Opus won't recognize what appears to be its latest message when asked if it said it, and that pasting those apparent messages can trigger a reasoning extraction classifier.
Combined views
8.1K
2 Sources, first seen ago
