Issue Station
Sign In
Tools hereToolbox

claude-code - 💡(How to fix) Fix Agent Teams via Remote Control: teammate permission prompts never render, causing silent deadlock [1 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
anthropics/claude-code#48549•Fetched 2026-04-16 06:57:03
View on GitHub
Comments
1
Participants
2
Timeline
5
Reactions
0
Author
magnum6actualmagnum6actual
Participants
github-actions[bot]github-actions[bot]magnum6actualmagnum6actual
Timeline (top)
labeled ×4commented ×1

Root Cause

  • The permission prompt is never shown to the user
  • The teammate silently blocks waiting for approval
  • The controller (team lead) sees only idle notifications — indistinguishable from normal multi-turn processing
  • The controller follows its idle patience protocol (wait, wait, nudge, escalate) but the nudge also gets no response because the teammate is blocked on the unrendered permission prompt
  • The build deadlocks with no visible indication of the cause

Fix Action

Workaround

Pre-approve common operations in permission settings (settings.json allowlist) before spawning teammates to reduce the likelihood of permission prompts. However, this doesn't eliminate the risk for unanticipated permission requests.

RAW_BUFFERClick to expand / collapse

Bug Report

When running Agent Teams via Claude Remote Control (Claude Desktop connected to a remote Claude Code instance), teammate permission prompts are never rendered in the Claude Desktop interface. The teammate silently blocks, the controller sees only idle notifications, and the build deadlocks.

Steps to Reproduce

  1. Enable Agent Teams (CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1) on a remote Claude Code instance
  2. Connect to it via Claude Desktop Remote Control
  3. Have the team lead spawn a teammate
  4. Have the teammate attempt an action that requires permission approval (e.g., writing to a path outside the project working directory)

Expected Behavior

The permission prompt should be rendered in the Claude Desktop Remote Control interface, allowing the user to approve or deny the action.

Actual Behavior

  • The permission prompt is never shown to the user
  • The teammate silently blocks waiting for approval
  • The controller (team lead) sees only idle notifications — indistinguishable from normal multi-turn processing
  • The controller follows its idle patience protocol (wait, wait, nudge, escalate) but the nudge also gets no response because the teammate is blocked on the unrendered permission prompt
  • The build deadlocks with no visible indication of the cause

Impact

This is a silent deadlock with no user-visible signal. The only way to diagnose it is to SSH into the remote host and inspect the teammate's process state directly. In autonomous workflows where the team lead runs multiple states without sponsor intervention, a single unrendered permission prompt can hang the entire build indefinitely.

Workaround

Pre-approve common operations in permission settings (settings.json allowlist) before spawning teammates to reduce the likelihood of permission prompts. However, this doesn't eliminate the risk for unanticipated permission requests.

Environment

  • Claude Code with CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1
  • Connected via Claude Desktop Remote Control
  • Remote host: Linux dev VM
  • Teammates spawned via Task tool with team_name parameter

extent analysis

TL;DR

The most likely fix involves modifying the Claude Desktop Remote Control interface to properly render teammate permission prompts, potentially by updating the handling of permission requests from teammates.

Guidance

  • Review the Claude Desktop Remote Control code to ensure it correctly handles permission prompts from teammates, possibly by checking for updates or modifications to the permission request handling logic.
  • Verify that the settings.json allowlist is properly configured to pre-approve common operations, reducing the likelihood of permission prompts.
  • Investigate the teammate's process state directly on the remote host to better understand the deadlock cause and identify potential fixes.
  • Consider implementing additional logging or monitoring to detect and diagnose silent deadlocks caused by unrendered permission prompts.

Example

No specific code snippet can be provided without more information on the Claude Desktop Remote Control interface code. However, the fix might involve updating a function or method responsible for rendering permission prompts, such as a renderPermissionPrompt function.

Notes

The provided information suggests that the issue is specific to the Claude Desktop Remote Control interface and the experimental Agent Teams feature. The fix may require updates to the interface code or changes to the permission request handling logic.

Recommendation

Apply a workaround by pre-approving common operations in the settings.json allowlist to reduce the likelihood of permission prompts, while also investigating and addressing the underlying issue with rendering permission prompts in the Claude Desktop Remote Control interface. This approach can help mitigate the risk of silent deadlocks until a full fix is available.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#installation#tensor shape#autograd error#model save/load

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING