What this error means

v2.1.141 regression: VSCode stream stalls at ~670 bytes after tool_use; UI hangs in thinking 5 min, work silently lost is a Claude Code failure pattern reported for developers trying to fix claude code vscode stream stalling after tool_use block, ui stuck in thinking state. Based on the imported evidence, treat this as a tool-specific troubleshooting page rather than a generic API error.

Why this happens

SSE stream sends ~650-700 bytes (tool_use block dispatched), then goes silent. Files modified server-side but assistant output lost. 5-minute idle timeout. Regression in v2.1.141.

Common causes

  • After upgrading Claude Code VSCode extension to v2.1.141, streaming responses die mid-turn at ~650-700 bytes. SSE stream stalls, webview shows 'thinking' for 5 minutes until timeout. Server-side tool_use already modified files on disk, but work is silently lost. Data loss regression in paid tool.
  • SSE stream sends ~650-700 bytes (tool_use block dispatched), then goes silent. Files modified server-side but assistant output lost. 5-minute idle timeout. Regression in v2.1.141.

Quick fixes

  1. Confirm the exact error signature matches v2.1.141 regression: VSCode stream stalls at ~670 bytes after tool_use; UI hangs in thinking 5 min, work silently lost.
  2. Check the Claude Code account, local tool state, and provider configuration involved in the failing workflow.
  3. Compare the failing environment with a known working setup, then change one configuration value at a time.

Platform/tool-specific checks

  • Verify the command, editor, extension, or API client that produced the error.
  • Compare local settings with CI, deployment, or editor-level settings when the error appears in only one environment.
  • Avoid deleting credentials, local model data, or project settings until the failing scope is clear.

Step-by-step troubleshooting

  1. Capture the exact error message and the command, editor action, or request that triggered it.
  2. Check whether the failure is account/auth, quota/rate, model/provider, local runtime, or deployment configuration.
  3. Review the source evidence below and compare it with your environment.
  4. Apply one change at a time and rerun the smallest failing action.
  5. Keep the working fix documented for the team or deployment environment.

How to prevent it

  • Keep provider/tool configuration documented.
  • Record non-secret diagnostics such as tool version, provider name, model name, and command path.
  • Add a lightweight check before CI or production workflows depend on the tool.