One of the most dangerous signals in Claude Code:


Determining 6 minutes—high effort—hundreds of tokens.
This usually means the model is stuck in a contextual quagmire. After that, it’s easy to produce off-target answers, mix up old memories, hallucinate the path, and misread tool outputs.
Don’t wait for it to “finish thinking.”
Stop directly, start a new session, and rerun with a clean context.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned