[BUG] Thinking blocks not expandable/collapsible in VS Code extension (renders as inert stub)

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

When using the Claude Code VS Code extension's graphical chat panel, extended thinking blocks render as a static "Thought for Xs" label with no way to expand them. Clicking the block does nothing — no expand arrow, no cursor change, no reaction. This happens consistently across every session and every prompt that triggers extended thinking, not just occasionally. It affects my day-to-day workflow directly, since I rely on reviewing Claude's reasoning while working across multiple related coding sessions, and right now I have no visibility into that reasoning at all in the graphical panel.

What Should Happen?

The "Thought for Xs" block should be clickable, expanding to show Claude's actual reasoning/summary text — the same way it works with older models (e.g. Sonnet 4.6), where clicking the block or pressing Ctrl+O reveals the full reasoning inline.

Error Messages/Logs

No error message or exception appears — the block simply renders as inert. There's nothing in the UI indicating a failure; it just silently shows no content.

Steps to Reproduce

1. Open a new session in the Claude Code VS Code extension's graphical chat panel (not Terminal mode).
2. Open the command menu (/) and confirm Extended Thinking is toggled on.
3. Switch model to [your model — e.g. Sonnet 5 / Opus 4.7 / Fable 5].
4. Send any prompt likely to trigger reasoning, e.g. "think through how to refactor this function step by step."
5. Wait for the response to fully render.
6. Observe the "Thought for Xs" label above the response — attempt to click it, or press Ctrl+O.
7. Nothing happens: no expand, no content shown, no visual feedback of any kind.
8. As a comparison: switching the same panel to Sonnet 4.6 and repeating steps 3–6 produces a normal, expandable thinking block — confirming the issue is tied to the newer model, not the panel itself.

Claude Model

Sonnet (default)

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.209

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Other

Additional Information

Additional context: this appears to be the same root cause reported in issues #49268, #49757, #66887, and #74910 — newer models default thinking.display to "omitted," and the extension isn't explicitly requesting display: "summarized" to override it.
using the VS Code extension's graphical chat panel, not a terminal (issue is specific to this panel and does not involve any terminal)