[BUG] Thinking mode toggle with tab button isn't working in 2.0.67

It sucks I know. But, try /config to switch Thinking off while waiting for them to bring this toggle key back.

At least "Thinking Off" should be default, it is more sensible IMO.

reference: https://github.com/anthropics/claude-code/commit/2192c86c208b39aa262ada534d89f72f65708a8e

That sucks, should at least introduce a new shortcut for that

Indeed, it's quite annoying to have to go to /config since it takes a bit of time to navigate and you can't switch to that mid-work, so you have to interrupt the AI first.

Please bring this feature back!!

I keep Thinking mode disabled by default due to the 5-hour thinking budget limit. When I'm not satisfied with Claude's output, I turn it on to get better reasoning. Typing /config adds 7 unnecessary keystrokes per switch.

But the real blocker is that I can't toggle Thinking or change models while editing. I have to either plan ahead or discard what I'm typing. Same issue with model switching.

Claude is capable enough that neither Thinking nor Opus are needed for every task. For simple tasks that need speed, I use Haiku. The point of shortcuts is to enable dynamic switching and optimize token usage. Without them, the optimization doesn't work. (Of course, it'd be ideal if Thinking and Opus's intelligence, and Haiku's speed were all cheap enough to use freely.)

Same issue with model switching.

You can use Alt+p to switch models while editing

+1. Please add an alternative shortcut for thinking mode toggle. Going through /config breaks the workflow.

I also found this problem today. Will have to return it. Ideally they have to make customizable shortcuts for popular(or any) action

I accidentally hit this Tab toggle all the time, so the motivation for the removal is understandable. However, the setting for the default isn't actually respected! I have to go to /config every time I start a conversation...

Also, an alt keyboard shortcut for this would be nice!

Discovered this problem today, because Claude kept thinking and chewing up more tokens. Couldn't turn it off with TAB. Discovered it's in /config but it's more annoying to get to AND it doesn't persist across sessions. Will switch back to enabled on session start. I really loved the TAB shortcut for thinking mode. I hope this is a bug and not a "feature" to remove it.

No! Muh tab!

Why break existing <TAB> behavior? Couldn't you just have used Ctrl+Space for prompt completion? It at least has history of being an "auto complete" shortcut.

Thanks for the reports. We removed the Tab shortcut because it was causing frequent accidental toggles. In our testing, thinking mode both reduces overall token usage and improves results, so we'd recommend keeping it on for most workflows.

We're aware the settings persistence is broken, a fix is coming in the next release.

Thanks for the reports. We removed the Tab shortcut because it was causing frequent accidental toggles. In our testing, thinking mode both reduces overall token usage and improves results, so we'd recommend keeping it on for most workflows. We're aware the settings persistence is broken, a fix is coming in the next release.

Thank you for replying.
Tab wasn't the right shortcut indeed. The solution to accidental toggles was to change it or allow to configure it though, not remove it entirely.

I personally find that thinking mode slows down Claude a lot, and in frequent cases seems to make it ramble incoherently with chains of "oh but wait" contradictions. Also want to disable it when I'm reaching end of context window and need quick completion. It's a case by case tool for me.

The statement that thinking reduces token usage is interesting. User's should be informed clearly if that's the case, with supporting data, because it's not intuitive.

i agree, half the time i'm working on a simple script that doesn't need thinking at all and jut makes it take longer than needed so being able to quickly toggle off and on is critical. I don't want to exit and change it every time. Even a slash command (like /model) isn't as user friendly IMO.

Given the issues you were seeing, switching the keypress for toggling thinking mode seems reasonable (although it could've been made more obvious that it had changed).

I've been toggling thinking mode quite often because it seems to use my quota rapidly. I reserve it for when I'm in planning mode on something non-trivial and turn it off to execute the plan. I like to see Claude's reasoning. Sometimes the thinking output mentions things I hadn't noticed and which it doesn't think are important, but which turn out to be useful. That can help me guide it when it's not thinking about something the right way or help me realize there's a better way to get something done.

That said, it would be nice to have fewer controls, especially as Opus 4.5's level-of-effort setting becomes available. There are getting to be several controls whose interactions aren't obvious. I don't envy you having to find a way to explain the various controls and choose which to make easily accessible. In the meantime, it'd be nice to have a shortcut or /command to see and change the level without having to use the interactive menu.

thanks!

You can add this to ~/.claude/settings.json:

```json
{
"env": {
"MAX_THINKING_TOKENS": "0"
}
}

  This disables thinking mode completely even toggled in /config, but this is a temporary workaround for people who don't want thinking by default, if you want to toggle thinking again you need to remove this configuration. 

Thanks for the reports. We removed the Tab shortcut because it was causing frequent accidental toggles. In our testing, thinking mode both reduces overall token usage and improves results, so we'd recommend keeping it on for most workflows. We're aware the settings persistence is broken, a fix is coming in the next release.

Will there be a way to set our own shortcut for this? For trivial changes it really degrades the DX to have to use thinking mode.

Just run claude with
MAX_THINKING_TOKENS=0 claude
or create your own wrapper for it
fast-claude
and add it to your PATH so you can launch fast-claude from anywhere

I use Claude Code a lot, and I always kept thinking turned off - I try to think together with the model, throwing quick messaages back and forth. Having thinking on breaks that completely - Claude keeps starting internal monologues that sometimes last a minute, and I have to sit around and wait for it to produce a small, and maybe already irrelevant answer.

Maybe thinking is good if you're trying to get better one-shot responses, but that's not what Claude is about for me. I constantly steer, adjust, and iterate, and that doesn't work now. (the fact that the setting keeps resetting to thinking: on, doesn't help)

I'd be keen to see the implied data of thinking mode saving token usage in general. Perhaps in high-level or decision-intensive tasks, but simple code it's really tedious to be forced into Thinking mode without an easy way out i.e. shortcut.

But yes, thanks for the reply!

I'd be keen to see the implied data of thinking mode saving token usage in general. Perhaps in high-level or decision-intensive tasks, but simple code it's really tedious to be forced into Thinking mode without an easy way out i.e. shortcut. But yes, thanks for the reply!

I second that

shortcut doesnt have to be a [TAB]
Some cmd + t or a sequence like cmd + / -> t would be arguably just as ergonomic

This was working perfectly in the previous version - the Tab key toggle was simple and intuitive. The new update broke this functionality.

Beyond this issue, Claude Code has multiple bugs right now, and Opus as a model has gone from incredible to catastrophic in terms of quality.

I really hope the team addresses these issues soon.

We removed the Tab shortcut because it was causing frequent accidental toggles.

@bogini

IMHO, the correct solution to avoid accidental toggles is to improve the shortcut, not remove it entirely, forcing users to go through the settings to enable or disable the feature.

the hell — i never accidentally toggled it and don't understand why it should ever happen. Also, the feedback in CLI is so clear and any accidental toggle is so easy to revert with another tap on the tab that it's a non-issue. Who is ever going to be going through config to turn it off and on in daily flow.

I had reported this myself when I noticed it and, after another case recently where the mobile app had thinking set to on by default, I decided for myself that there's something more profit-driven at heart here (which is a real shame given Anthropic's stated mission).

Start here if what I'm saying seems to resonate, I'm keeping my own issue open for meta-discussion of potential deeper problems here.

I'd be keen to see the implied data of thinking mode saving token usage in general. Perhaps in high-level or decision-intensive tasks, but simple code it's really tedious to be forced into Thinking mode without an easy way out i.e. shortcut.

Thirded, this data should have been published and cited within the first reply IMO (or at least taken this issue as a way to quietly publish this if they didn't want to draw so much attention to it, but if it were true, I think they'd be making a big PR push to get everyone to turn thinking mode on _themselves_; I just think they might be worried the average user doesn't even know what thinking is and won't notice it getting set on or off/consuming their limits more quickly).

Can we please get a shortcut similar to alt + p (for switching models)? 😕

this bug seems to be back in 2.0.70

In version 2.0.70, the Thinking mode indicator is no longer visible in the input field. It should be displayed explicitly so users know when Thinking mode is active.

I'm struggling to make use of the current claude code due to this, especially since I rely on Opus almost exclusively.

My workaround is to keep using the older version:

npm install -g @anthropic-ai/claude-code@2.0.65

& disable auto-updates in ~/.claude/settings.json

{
   "env": {
      "DISABLE_AUTOUPDATER": "1"
   }
}

Hopefully they reintroduce some shortcut so I don't have to use this older version.

+1 one to bring this back, Caude is responding faster with no thinking which is highly desired for certain tasks, questions, simple queries, having to use config or running a separate instance only for this degrades the UX

Thanks for the reports. We removed the Tab shortcut because it was causing frequent accidental toggles.

@bogini you realize how dishonest this seems when the team has immediately switched to using that hotkey for another feature which can have undesirable results if accidentally triggered right? It makes it sound like a retroactive justification rather than the actual reasoning.

Now of course, seeming dishonest does not make it dishonest. But the optics aren't great here.

PSA - they removed the indicator in the input box + the "Thinking..." status got merged into the normal indicator. You can still turn it off via /config, and it seems to persist (at least it didn't get reset after restarting claude).

@bogini this looks like you guys are doing Apple-style design decisions here - "we know better what our users need". I don't know what metrics you have that made you think it's a good idea, but consider this please.

"Thinking" is basically generating its own prompt after the user's prompt and trying to follow both. It makes sense that it improves results for one-shot prompts / people who like to leave claude unsupervised / people who want to build prototypes / just tell claude about the bug without understanding what's going on in the code - actual vide coding. I don't hate those people, but IMO they're in the dead branch of LLM-assisted development - you can't produce reliable software that way, and I'm pretty sure that Anthropic staff have tried and know that too.

What works, is to steer claude continually. Instead of asking it to do something and hoping that works, I:

1. ask it to do something
2. ask for the implementation plan
3. review the plan - it usually involves talking to it and telling it lots of extra context I didn't realize it needed
4. _critically_, monitor every code change change it makes and steer (Esc > tell it what's wrong) until it's done

Basically, that's pair programming with a middle-level developer who reads and types really fast. _This_ is what works, _this_ is the future, and you guys should be promoting it with your UI decisions.

Thinking mode is incompatible with this - it randomly stops for tens of seconds in the middle of the flow to self-steer.

Turn thinking off by default, promote plan mode. Maybe a separate mode for non-vibecoders that stops it rushing to make stuff before we both have an understanding of what it's building?

PSA - they removed the indicator in the input box + the "Thinking..." status got merged into the normal indicator. You can still turn it off via /config, and it seems to persist (at least it didn't get reset after restarting claude). @bogini this looks like you guys are doing Apple-style design decisions here - "we know better what our users need". I don't know what metrics you have that made you think it's a good idea, but consider this please. "Thinking" is basically generating its own prompt after the user's prompt and trying to follow both. It makes sense that it improves results for one-shot prompts / people who like to leave claude unsupervised / people who want to build prototypes / just tell claude about the bug without understanding what's going on in the code - actual vide coding. I don't hate those people, but IMO they're in the dead branch of LLM-assisted development - you can't produce reliable software that way, and I'm pretty sure that Anthropic staff have tried and know that too. What works, is to steer claude continually. Instead of asking it to do something and hoping that works, I: 1. ask it to do something 2. ask for the implementation plan 3. review the plan and do multiple iterations on it 4. _critically_, monitor every code change change it makes and steer (Esc > tell it what's wrong) until it's done Basically, that's pair programming with a middle-level developer who reads and types really fast. _This_ is what works, _this_ is the future, and you guys should be promoting it with your UI decisions. Thinking mode is incompatible with this - it randomly stops for tens of seconds in the middle of the flow to self-steer. Turn thinking off by default, promote plan mode. Maybe a separate mode for non-vibecoders that stops it rushing to make stuff before we both have an understanding of what it's building?

+1

I'm quoting your full reply partly to simply increase it's presence in all future reader's browsers because I agree with is so whole-heartedly. To type a thorough response to what you've laid out here would be practically quite duplicative, but I'll add that I'm trying to get people to think beyond the engineering staff and think about what their surrounding corporation is likely forcing upon them for profit motives which destroy the purity of what could be a very strong union between AI/VC investment and the wisdom of the old guard (and those inspired by them) who understand that a terminal is more than a little box in their IDE.

I fear, though, that the deep entaglement of this codebase with Anthropic's proprietary model will ultimately doom it to either a _really_ good fork or the meaningless and performative "Open Source" practices of companies who try to win the old guard's wisdom (and, more importantly, free labor) with their "principle" which tends to be poorly disguised capitalist tact.

Here's another entrypoint (just jumps to my reply above) which leads to my own issue centered more around the corruption of pure principles this change represents, I'd encourage others who've read this far to at least think for themselves about these issues and decide firmly where they stand before a company decides for them, because we should all know how their decisions will be biased.

+1 tab to toggle thinking mode

Temporary workaround: roll back Claude Code to version 2.0.65 and disable the auto updater.

From my experience using this every single day (not at all fact), Anthropic made Opus 4.5 'default', but decreased its token usage in certain parts of the pipeline so it didn't destroy your usage, like Opus 4.1 did (and I'm on the x20 max plan), with the perceived benefit for users being that they are using the powerful 4.5 model at a higher rate than a few weeks ago would allow with 4.1.

I personally have seen a decrease in performance slightly since 4.5 became the default.

I build software with many various model api's, and deploy them on AWS, and I've realized it's SO easy to tell users one thing on the face, and then decrease compute costs in any way you want in the back.

Users only know 'hey I'm using the fancy new model woohoo!'. You really never know what they're actually giving you across the series of steps engineered between your prompt, the model's response, and the management of long-context (among other areas).

While we wait in anger for this bug's resolution, it's also interesting to think about the business decisions behind a seemingly 'stupid' product update (and yes it's a stupid update).

Thanks for the feedback. We removed the Tab shortcut based on data showing high accidental toggle rates and performance data on thinking mode. We heard clearly the ask was for a less conflicting shortcut, not removal.

2.0.71 adds Alt+T to toggle thinking mode inline. Works while typing, similar to Alt+P for model switching. This toggle is session-scoped for quick adjustments. /config remains available for persistent settings across sessions.

We're constantly evaluating usage patterns to improve the experience for the majority of users. Sometimes that means making tradeoffs, and feedback like this helps us find the right balance. Thanks for pushing on this.

Thanks for the feedback. We removed the Tab shortcut based on data showing high accidental toggle rates and performance data on thinking mode. We heard clearly the ask was for a less conflicting shortcut, not removal. 2.0.71 adds Alt+T to toggle thinking mode inline. Works while typing, similar to Alt+P for model switching. This toggle is session-scoped for quick adjustments. /config remains available for persistent settings across sessions. We're constantly evaluating usage patterns to improve the experience for the majority of users. Sometimes that means making tradeoffs, and feedback like this helps us find the right balance. Thanks for pushing on this.

Amazing! Thank you for sharp feedback turnaround!

Thanks for the feedback. We removed the Tab shortcut based on data showing high accidental toggle rates and performance data on thinking mode. We heard clearly the ask was for a less conflicting shortcut, not removal. 2.0.71 adds Alt+T to toggle thinking mode inline. Works while typing, similar to Alt+P for model switching. This toggle is session-scoped for quick adjustments. /config remains available for persistent settings across sessions. We're constantly evaluating usage patterns to improve the experience for the majority of users. Sometimes that means making tradeoffs, and feedback like this helps us find the right balance. Thanks for pushing on this.

really helpful, thanks!

I thought my claude was just broken. Regardless of peoples personal opinions on whether or not thinking mode is useful, disabling it when it was already enabled on my end, then removing the normal keybind, made it seem like it was just removed from claude entirely. I guess I'm downgrading and killing updates? This is a massive UX shift for those of us who do use it.

This could have been solved with prompts and indicators. First time you enable thinking, "did you mean to do that?". User clicks no, you prompt them to switch to the altbind. They click yes, you don't bother them again. But to just strip it entirely, start removing indicators, and then make it per-session so I have to rewind and waste more tokens etc... eugh.

unfortunately alt+t doesn't work for me - I am on a macOS, using Polish keyboard layout, and GhosTTY (tried iTerm2 as well)

Same issue here. It also doesn't display thinking mode indicator (blue border) anymore when it's on.

Same on MacOS using a different keyboard layout. Alt + t is not working reliably. The lack of the thinking mode indicator is annoying, too.

You might needs to configure your terminal to use Option as Meta to Alt+t to work:

• Terminal.app: Settings → Profiles → Keyboard → Check "Use Option as Meta key"
• iTerm2: Settings → Profiles → Keys → General → Left Option Key → "Esc+" (Same for Right Option Key if desired)
• VS Code integrated terminal: Add to settings.json: "terminal.integrated.macOptionIsMeta": true

@bogini
Thanks for reintroducing a keybinding (at least it's usable now), but please consider not showing a prompt when using Alt + T.
We went from "Tab" (single key) to Alt + T + Up/Down arrow + Enter = 4 keys.
Seeing as it's a boolean, Alt + T alone should be sufficient to toggle between the two values - no prompt needed.

@bogini We went from "Tab" (single key) to Alt + T + Up/Down arrow + Enter = 4 keys.

@gajop-ptx Actually we actually went from typing the word “think” or “think harder” and so on, where we had granular control over this to a single thinking mode “tab” to be hidden after multiple keys.

I generally don’t comment to repeat the same as other users but the lack of easy thinking mode switch is a deal breaker for my workflow and will push me to consider alternatives if not fixed.

@andreazorzetto

@gajop-ptx Actually we actually went from typing the word “think” or “think harder” and so on, where we had granular control over this to a single thinking mode “tab” to be hidden after multiple keys.

I think that had its own problems, as you couldn't type a sentance like this without accidentally triggering thinking. It being controlled through a keybinding was an improvement in that regard imo...

For anyone using Ghostty, you need to add these key binding settings to your Ghostty config so both alt+p and alt+t work:

macos-option-as-alt = left
keybind = alt+left=unbind
keybind = alt+right=unbind

I only added the line keybind = alt+left=unbind. That way my left key is acting as Alt and my right key as Option.

@gajop-ptx

as you couldn't type a sentance like this without accidentally triggering thinking

Hmm, I definitely remember triggering thinking accidentally that way. But what about Help me refactor this component, !think about how we can reduce duplication (return the old system but only react to !think instead of think)?

Thanks for the feedback. We removed the Tab shortcut based on data showing high accidental toggle rates and performance data on thinking mode. We heard clearly the ask was for a less conflicting shortcut, not removal. 2.0.71 adds Alt+T to toggle thinking mode inline. Works while typing, similar to Alt+P for model switching. This toggle is session-scoped for quick adjustments. /config remains available for persistent settings across sessions. We're constantly evaluating usage patterns to improve the experience for the majority of users. Sometimes that means making tradeoffs, and feedback like this helps us find the right balance. Thanks for pushing on this.

Thanks a lot. thats way better then using /config. still not so good then simble Tab ... but better. So the bug that the visual feedback is gone is still ongoing, right? The text "Thinking on" is disapering after a sec. ist must stay there. Or fix the bug that the blue outlines are visible again, please.

As a workaround does the ultrathink word still triggers thinking mode?

why the hack they did this

why the hack they did this

I actually like their approach

TAB now does prompt-autocomplete just like a _shell_ should

Claude Code is primarily an AI _shell_, like bash but with ai features

The new shortcut (Alt + Tab) does not work properly on MacOS / kitty terminal.

It's very uncomfortable, I can't breathe on my Mac switching in Think mode

Can CaludeCode have a place to support user-defined shortcut configuration?

The current 'alt+t' or 'alt+p' under the ABC input method gives the characters '†' and 'π' 😳

This is a really frustrating update, as it altered a lot of existing workflows with zero warning.
Please consider reverting this, or at least allowing the key to be customized.

🚧 Workaround: Custom One-Press Thinking Mode Toggle

Run the prompt below and Claude Code will adopt a workaround itself on your machine.

[!TIP] - If it runs unreliably for you, just increase delays (they're razor-thin at the moment) - Ask Claude Code to give you options if you're on other OS than macOS - Recommended: Opus 4.5 model

https://github.com/user-attachments/assets/b6633860-8ea0-45f8-9dd4-5f3d7c2920bb

A macOS hotkey to instantly toggle Claude Code's thinking mode via skhd.

This skhd-based solution works system-wide, instantly, with zero terminal config.

---

1. Implementation Prompt for Claude Code

Use this prompt to have Claude Code set up the toggle on a fresh macOS machine:

Set up a macOS-wide hotkey (Option+T) that toggles Claude Code's "thinking mode" setting.

Requirements:
- Use skhd for the hotkey binding (install via homebrew if needed)
- Use AppleScript via osascript to send keystrokes to the focused window
- The script should work when any Claude Code terminal window is focused

The keystroke sequence to toggle thinking mode is:
1. Type "/config" (opens completion menu)
2. Press Enter (selects /config and opens config panel)
3. Press Down arrow twice (navigates to "Thinking mode" which is the 3rd item)
4. Press Enter (toggles the setting)
5. Press Escape (closes the panel)

Create this script at ~/scripts/toggle-thinking.sh:

#!/bin/bash
# Toggle thinking mode in Claude Code via keystrokes
# Triggered by skhd on Option+T

osascript <<'EOF'
tell application "System Events"
    delay 0.2   -- wait for Option key release
    keystroke "/config"
    delay 0.06  -- 3x safety margin over 0.02s minimum tested
    key code 36   -- Enter (opens config panel)
    key code 125  -- Down
    key code 125  -- Down (now on "Thinking mode")
    key code 36   -- Enter (toggle)
    key code 53   -- Escape (close)
end tell
EOF

Then:
1. Make the script executable
2. Install skhd if not present: brew install koekeishiya/formulae/skhd
3. Start skhd service: skhd --start-service (or brew services start skhd)
4. Add to ~/.skhdrc (create if needed): alt - t : ~/scripts/toggle-thinking.sh
5. Reload skhd: skhd --reload

Test by focusing a Claude Code window and pressing Option+T.

---

2. How This Was Built

Claude Code developed this through iterative testing with direct terminal feedback:

1. Initial approach: Created an AppleScript that sends keystrokes to toggle the config. First attempt bundled /config with Enter in a single send - this caused timing issues.

1. Debugging via terminal observation: By launching a Claude Code instance in a kitty terminal and using kitty @ get-text to read the screen contents after each keystroke, Claude could see exactly what was happening:

• Discovered the completion menu needs time to render
• Found that arrow key codes (125 for down) work correctly with AppleScript
• Identified the config panel opens with a single Enter, not two

1. Iterative refinement: Sent keystrokes one at a time, checking screen state after each:

``
After /config: completion menu visible
After Enter: config panel open, cursor on "Auto-compact"
After Down: cursor on "Show tips"
After Down: cursor on "Thinking mode"
After Enter: setting toggled
After Escape: panel closed, log shows "Enabled/Disabled thinking mode"
``

1. Delay optimization: Systematically tested each delay to find failure points:

• Started with conservative delays totaling ~0.5s
• Tested each delay individually, reducing until toggle failed
• Findings:
• delay after /config: 0.02s minimum (0.01s fails ~50% of time)
• delay after Enter: 0s (no delay needed)
• delay between Downs: 0s (keys buffer correctly)
• delay after toggle Enter: 0s (no delay needed)
• Final script: single 0.06s delay (3x safety margin) - still instant

Total development time: ~20 minutes of iterative testing with terminal feedback.

---

3. Message to Claude Code Maintainers

API Access for Config Management is Critically Needed

Thank you for responding to #13759 and adding Alt+T in 2.0.71. However, the implementation still has gaps:

1. Terminal dependency: Requires Option as Meta config per-terminal
2. Keyboard layout issues: Doesn't work with non-US layouts (Polish, etc.)
3. Not a true toggle: Opens a menu requiring additional keystrokes
4. No programmatic access: Still can't script config changes

The fact that users need to resort to simulating keystrokes via AppleScript to get a reliable toggle demonstrates a deeper gap in Claude Code's extensibility.

What's needed:

# Dream API - read config by session
claude config get thinking_mode --session <session-id>

# Dream API - write config by session
claude config set thinking_mode true --session <session-id>

# Or via environment/socket
CLAUDE_SESSION=<id> claude config set thinking_mode true

Why this matters:

1. Workflow automation: Power users want to script their Claude interactions. Toggling thinking mode, changing models, adjusting verbosity - these should be one command, not a fragile keystroke simulation.

1. IDE integrations: VS Code extensions, Neovim plugins, and other integrations can't properly control Claude Code instances without programmatic config access.

1. Multi-session management: Users running multiple Claude sessions need per-session control. Currently there's no way to target a specific instance.

1. Reliability: AppleScript keystroke simulation breaks when:

• UI changes (menu item order)
• Timing varies (slower machines need longer delays)
• Focus is lost mid-sequence
• Terminal emulators handle key codes differently

1. Community sentiment: Search Discord/GitHub - config automation is a recurring request. The current answer of "open the config panel manually" frustrates power users who chose a CLI tool specifically to avoid manual UI interactions.

Minimum viable solution:

• Expose a Unix socket per session (like kitty does)
• Accept JSON-RPC commands for config read/write
• Document the protocol so the community can build tooling

This would transform Claude Code from "a CLI tool" to "a programmable AI coding assistant" - which is what your power users actually want.

---

Created by Claude Code through iterative terminal-based development and testing.

@Elijas

SOLVED: Claude Code Thinking Mode Toggle (Option+T)

Is this a joke? A breaking change is introduced, no notice, no semantic versioning, and paying users need to spend time messing with the terminal hoping a shortcut would get everything to work as before? Doesn't sound like a reason to celebrate

@Elijas > SOLVED: Claude Code Thinking Mode Toggle (Option+T) Is this a joke? A breaking change is introduced, no notice, no semantic versioning, and paying users need to spend time messing with the terminal hoping a shortcut would get everything to work as before? Doesn't sound like a reason to celebrate

😆 fair

I propose we keep pushing for a real fix while using workarounds to stay productive

what do you think @andreazorzetto ?

Sometimes I really question the sanity of some decisions by the Claude code team here.
Did anyone actually complain about accidental thinking mode? NO!
But instead to open a issues and let users comment, lets push some SHIT out which breaks users workflows!
For me on a mac the new shortcuts dont even work. 💩

If you people want to get this changed back, write to Thariq and complain https://x.com/trq212

It's very uncomfortable, I can't breathe on my Mac switching in Think mode Can CaludeCode have a place to support user-defined shortcut configuration? The current 'alt+t' or 'alt+p' under the ABC input method gives the characters '†' and 'π' 😳

Fix comming:
https://github.com/anthropics/claude-code/issues/14327#issuecomment-3671296047

this is totally uncomfortable!! Using non-US keyboard is having a hard time to toggle thinking mode.
please fix this soon!

Really unacceptable that this is still an issue.

Still an issue January 6 😠

Fix it!

jesus christ fix this shit how is this even still open

i'm on macbook M4 pro , using shift+tab i was able to switch for thinking mode . If any one still looking for a solutiont

This was an intentional keybinding change in v2.0.72 — the thinking toggle moved from Tab to Alt+T to prevent accidental triggers when indenting text or using tab completion.

From the changelog:

Changed thinking toggle from Tab to Alt+T to avoid accidental triggers

If you prefer Tab, you can remap it in ~/.claude/keybindings.json:
~~~json
{
"thinking:toggle": "tab"
}
~~~

Closing as by-design.

This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.