Multi-Machine Claude Sessions via Shared Filesystem

Preflight Checklist

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

Problem Statement

When working on a single codebase synced across multiple machines (e.g., via Resilio Sync, Syncthing, or Dropbox), developers can only use Claude Code on one machine at a time. There's no way to coordinate parallel Claude Code sessions across machines working on the same project — for example, having Claude edit backend files on a desktop while simultaneously editing frontend files on a laptop. This forces sequential workflows on projects that would naturally benefit from parallel development, especially for solo developers who maintain multiple workstations.

Proposed Solution

Allow Claude Code sessions to coordinate across machines via a shared filesystem. A lightweight lock/manifest file (e.g., .claude/multi-session.json) in the project root would let multiple Claude Code instances discover each other and divide work. Each session would claim specific files or directories, preventing conflicts while enabling true parallel edits. The coordinating session would act as the "orchestrator," dispatching tasks to connected sessions and merging results. This could leverage existing file sync tools (Resilio Sync, Syncthing, Dropbox) as the transport layer rather than requiring Anthropic to build new networking infrastructure.

Alternative Solutions

1. SSH-based remote editing — Claude Code could SSH into a second machine and edit files directly, but this requires network configuration, open ports, and doesn't work well across NATs or firewalls.

1. Git-based workflow — Work on separate branches on each machine and merge afterward, but this loses the real-time coordination benefit and requires manual conflict resolution.

1. Single machine with more resources — Instead of parallelizing across machines, use a single powerful workstation, but this doesn't help developers who already have multiple machines and want to utilize idle compute.

1. Claude Desktop Cowork on one machine — Use the autonomous Cowork feature for background tasks, but this is limited to a single machine and doesn't leverage available hardware on other workstations.

Priority

High - Significant impact on productivity

Feature Category

Other

Use Case Example

A developer works from home on a laptop and has a desktop at the office, with project files kept in sync via Dropbox. They need to add a photo gallery page and redesign the navigation menu — two independent tasks affecting separate files. Currently, they must complete one task on whichever machine they're at, wait for the files to sync to the other machine, then start the second task next time they're at that location. With multi-machine coordination, they could instruct Claude on the home laptop to build the photo gallery while simultaneously having Claude on the office desktop redesign the navigation — both machines working in parallel regardless of where the developer physically is.

Additional Context

This feature would complement the existing Claude Desktop Cowork feature by extending autonomous capabilities across multiple physical machines. The coordination mechanism could be filesystem-based (shared lock/manifest files), requiring no new networking infrastructure — just leveraging existing file sync tools that many developers already use.