claude-code - 💡(How to fix) Fix [FEATURE] iOS: start new Claude Code sessions on my paired Mac, with full feature parity [1 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#57904Fetched 2026-05-11 03:22:23
View on GitHub
Comments
0
Participants
1
Timeline
3
Reactions
0
Author
Participants
Timeline (top)
labeled ×2renamed ×1

Root Cause

That blocks an entire class of workflows. The session needs to run on my actual Mac — with its logged-in browser, its full filesystem, its locally-configured MCPs, and its native desktop apps — because that's where all my work, credentials, and tools live. Nothing in the cloud has access to those, and nothing in the cloud can be made to.

Fix Action

Fix / Workaround

  • #48949 (Persistent always-on Remote Control) — about auto-enabling RC on already-started sessions. Doesn't address starting new sessions from mobile.

  • #46460 (Use Dispatch from Claude Desktop) — about cross-workstation routing, different surface.

  • #15922 / #25115 (Mobile Companion App) — about push notifications and approve/deny on existing sessions. Steering, not starting.

  • Start a session at the desk before leaving. Today's workaround. Doesn't work when the user isn't at the desk in the first place.

  • SSH from mobile. Functionally possible, lacks the native UX (push notifications, diff approval, mobile-rendered session view).

RAW_BUFFERClick to expand / collapse

Preflight Checklist

  • I have searched existing requests and this feature hasn't been requested yet
  • This is a single feature request (not multiple features)

Problem Statement

Today, the only way to start a new Claude Code session is from my Mac — by opening a terminal and running claude. Once a session is running, I can pair it to mobile with /remote-control and steer it from my phone. But if I'm not physically at my Mac, I can't start a new session at all.

That blocks an entire class of workflows. The session needs to run on my actual Mac — with its logged-in browser, its full filesystem, its locally-configured MCPs, and its native desktop apps — because that's where all my work, credentials, and tools live. Nothing in the cloud has access to those, and nothing in the cloud can be made to.

The asymmetry today:

  • From my desk: I can start a Claude Code session on my Mac with full local capabilities. ✅
  • From my phone: I can steer sessions I already started at my desk. I cannot start a new one. ❌

The ask is to close that asymmetry.

Proposed Solution

Let the iOS app start new Claude Code sessions on the user's paired Mac, with full Claude Code feature parity — equivalent to having run claude from a terminal on that Mac.

The new session should inherit everything a desk-started session would:

  • Full filesystem access at a user-specified cwd
  • Logged-in local browser
  • All locally-configured MCPs (computer-use, claude-in-chrome, scheduled-tasks, custom connectors with host credentials)
  • Native computer use against the desktop
  • The user's real shell, env, and credentials
  • All custom agents, skills, slash commands, hooks, and settings from ~/.claude/
  • Support for multiple concurrent sessions, not one-at-a-time

Once started, the session should appear in the iOS app exactly as if it had been launched from the desk and /remote-control-paired — same UI, same controls, same lifecycle.

Why existing issues don't cover this

  • #48949 (Persistent always-on Remote Control) — about auto-enabling RC on already-started sessions. Doesn't address starting new sessions from mobile.
  • #46460 (Use Dispatch from Claude Desktop) — about cross-workstation routing, different surface.
  • #15922 / #25115 (Mobile Companion App) — about push notifications and approve/deny on existing sessions. Steering, not starting.

Spawn ≠ steer. Worth its own thread.

Alternative Solutions

  • Start a session at the desk before leaving. Today's workaround. Doesn't work when the user isn't at the desk in the first place.
  • SSH from mobile. Functionally possible, lacks the native UX (push notifications, diff approval, mobile-rendered session view).

Priority

High — Significant impact on productivity for any workflow whose value depends on logged-in local state, native desktop apps, computer use against the actual desktop, locally-configured MCPs, or files outside any GitHub repo.

Feature Category

iOS app / Remote Control

Use Case Example

I'm out and want Claude to audit a few of my logged-in browser tabs for outstanding items, summarize what's actionable, and queue follow-ups in my local notes system. This requires my actual Chrome, my actual local files, and my actual MCPs — none of which are accessible to anything except a session running on my Mac.

Today: blocked until I'm home.

With this feature: from the iOS app, start a new Claude Code session on my Mac. It spawns with full local capabilities, and I drive it from my phone.

The same pattern applies to any mobile-initiated workflow touching local notes, screenshots, native desktop apps, logged-in web services, or any MCP that requires host credentials.

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…

Still need to ship something?

×6

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

Back to top recommendations

TRENDING

claude-code - 💡(How to fix) Fix [FEATURE] iOS: start new Claude Code sessions on my paired Mac, with full feature parity [1 participants]