anthropics / anthropics/claude-code

[BUG] Cowork: every follow-up prompt hangs on "Working through a complex response…" — reproduces on desktop AND claude.ai web

Ouverte
#91,004 1 commentaire 0 réactions 0 personnes assignées Voir sur GitHub
bug
Langage dominant
Python
Étoiles
145k
Forks
23.1k
Métriques de merge des PR
Métriques de PR en attente

Description

### Preflight Checklist

- [x] I have searched [existing issues](https://github.com/anthropics/claude-code/issues?q=is%3Aissue%20state%3Aopen%20label%3Abug) 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?

[BUG] Cowork: every follow-up prompt hangs on "Working through a
complex response…" — reproduces on desktop AND claude.ai web

Onset: first observed 2026-08-30, first use after a two-day break.

Behavior:
- First prompt in a session completes normally
- Every subsequent prompt hangs indefinitely on "Working through a
complex response…"
- Cancel button is non-functional — the session cannot be stopped
- Only recovery is a new session, which then fails identically on its
second prompt

Scope: 5 separate Cowork projects, identical behavior in all.

Reproduces in BOTH execution environments:
- Desktop app 1.40609.0 (f65e38), build 2026-08-27T16:19:51Z, macOS
Tahoe 26.6.2 (local VM)
- claude.ai in browser (cloud sandbox)

Since these are separate sandboxes with different hardware, kernel and
network paths, the failure appears to sit in shared session
orchestration rather than in either execution environment.

Ruled out:
- All MCP servers / connectors disabled — still hangs
- All skills disabled — still hangs
- No plugins enabled
- Not project-specific
- macOS 26.6.2 installed 2026-08-26, before a Cowork session that ran
normally on 2026-08-27

Not a duplicate of #45531: there the local VM never initializes. Here
sessions start and the first prompt completes.

Plan: Pro
Suggested label: area:cowork

### What Should Happen?

Claude should allow follwup prompts to work without hanging

### Error Messages/Logs

```shell

```

### Steps to Reproduce

Behavior:
- First prompt in a session completes normally
- Every subsequent prompt hangs indefinitely on "Working through a
complex response…"
- Cancel button is non-functional — the session cannot be stopped
- Only recovery is a new session, which then fails identically on its
second prompt

### Claude Model

Not sure / Multiple models

### Is this a regression?

Yes, this worked in a previous version

### Last Working Version

_No response_

### Claude Code Version

Claude 1.40609.0 (f65e38) 2026-08-27T16:19:51.000Z

### Platform

Anthropic API

### Operating System

macOS

### Terminal/Shell

Terminal.app (macOS)

### Additional Information

_No response_

Guide de contribution

Aucun guide de contribution indexé pour ce dépôt

Piste de recherche

No repository file, test, log, or entry point is named. Start by reproducing the second-prompt hang in both listed environments and comparing the shared session behavior, then inspect the session orchestration path and capture relevant logs. Done means follow-up prompts complete and Cancel stops the session without requiring a new session.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Évaluation

Stack technique
macos, python
Domaine
api, backend
Type d'issue
Bug
Difficulté
5/5
Temps estimé
Plus d'une semaine
Activité
Active
Clarté
À clarifier
Accessibilité débutants
35/100

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.