AgentArmorHQ
Sign in
  • AgentArmorHQ teamTeam4d agoNeeds helpSolved

    How do I see Claude Code's thinking without pressing Ctrl+O and Ctrl+E after v2.0?

    After upgrading Claude Code to v2.0.0, the model's thinking no longer appears in verbose mode by default. Users who want to follow the reasoning must press Ctrl+O, then Ctrl+E, and scroll up each time. Is there a setting to always show thinking, or a way to get the earlier behavior back?

    Originally asked on GitHub: https://github.com/anthropics/claude-code/issues/8477

    4 answers

    • Best answerAgentArmorHQ teamTeam4d ago

      The issue is marked closed as completed, but the thread shows no maintainer reply or fix that restores thinking by default. Treat the status as open.

      Workaround from the thread: press Ctrl+O to open verbose mode, then Ctrl+E to expand thinking, and scroll up. The reporter calls this tedious. Another commenter says it is too late to be useful for live review, because the thinking only appears after Claude finishes.

      Other workaround: install the older release with npm i -g @anthropic-ai/claude-code@1.0.128. A commenter says this version displays thinking, but it means giving up v2.0 features. Another commenter notes this only makes sense if you don't rely on the newer features.

      A commenter believed the alwaysThinkingEnabled setting turns on thinking mode at startup, so you don't have to press Tab. That commenter was describing what they believed the setting was meant for, not testing it, and it was not stated as controlling whether thinking is visible. No maintainer confirmed this. Check its behavior in your version before relying on it.

      Did this work for you?
    • AgentArmorHQ teamTeam4d ago

      What a GitHub user tried on GitHub (Partly worked)

      The reporter pressed Ctrl+O to open verbose mode, then Ctrl+E to expand thinking, and scrolled up to read it each time.

      Source: https://github.com/anthropics/claude-code/issues/8477 (opens external site)

      Did this work for you?
    • AgentArmorHQ teamTeam4d ago

      What @davidjson tried on GitHub (Outcome unclear)

      Suggested downgrading with npm i -g @anthropic-ai/claude-code@1.0.128 to show thinking again, for users who do not depend on newer features.

      Source: https://github.com/anthropics/claude-code/issues/8477#issuecomment-3361642642 (opens external site)

      Did this work for you?
    • AgentArmorHQ teamTeam4d ago

      What @anthrotype tried on GitHub (Outcome unclear)

      Believed the alwaysThinkingEnabled setting enables thinking mode at startup so users don't need to press Tab, rather than controlling visibility. Not tested in the thread.

      Source: https://github.com/anthropics/claude-code/issues/8477#issuecomment-3486969155 (opens external site)

      Did this work for you?

    Add your answer or comment