[BUG] Tool calls frequently break

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?

Recently, Claude has been failing to call tools very frequently, regardless of the effort level.
Read, Write, AskUserQuestion…anything. It even feels like it's getting worse day by day.

I've been using a temporary, symptomatic workaround I recently saw in a technical article to prevent it from simply breaking and stopping, but that's not a fundamental solution.
Anthropique should confront the fact that this is wasting tokens and a huge amount of time.
It's wasting your limited computing resources.

What Should Happen?

Claude should be able to use tools correctly.

Error Messages/Logs

court

  [{"question": "CR-SA-17-E3 検証完了（Closed 判定）後の次の進め方は？（選択肢以外の入力も可能です）", "header": "次アクション", "multiSelect": false, "options": [{"label": "E4 を発行して継続", "description":
  "Scene 3 進行統合 = 系列完遂。ペース実挿入 + 可変構成走行 + fixValue ギャップ対応。完遂で push 判断へ"}, {"label": "ここで区切る", "description": "PM ドキュメント整備（STATE 書き直し +
  作業ログ）してセッション終了。E4 は次セッションで"}, {"label": "E1〜E3 を先に push", "description": "画面変化が出る E3 まで公開。ただし可変構成の実走行は E4 未完のため中間状態に注意"}, {"label":
  "Claudeに任せる", "description": "系列完遂までの距離とコンテキスト状況を見て最適な次アクションを選ぶ"}]}]

Steps to Reproduce

This isn't a problem that only occurs through specific procedures.
It can happen suddenly at any time, during any task.

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

2.1.168

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Windows Terminal

Additional Information

When I point out a mistake, or when it's automatically pointed out by a Hook, Claude responds as if he's aware that the cause is a syntax error in his tool call.
I had a tool-call syntax error. Let me reissue the approval question correctly.