
Memory Sync
- Updated July 28, 2026
- cgraf78/claude-plugins
A productivity skill that syncs Claude Code project memory directories to a cloud drive so context is shared across machines. It keeps agent memory consistent for multi-device workflows. Useful for developers working from more than one computer.
Key points
- Memory directory sync
- Cloud drive
- Cross-machine sharing
Memory Sync by the numbers
- Data as of Jul 29, 2026 (Skillselion catalog sync)
/plugin marketplace add cgraf78/claude-plugins/plugin install memory-sync@cgraf78-claude-pluginsAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Last updated | July 28, 2026 |
|---|---|
| Repository | cgraf78/claude-plugins ↗ |
What it does
Sync Claude Code project memory directories to a cloud drive to share context across machines.
README.md
memory-sync
Syncs Claude Code project memory directories to a cloud drive, so memory is shared
across machines even when the home directory path differs (e.g. /Users/chris vs /home/chris).
Each ~/.claude/projects/*/memory/ directory is replaced with a symlink pointing to a
consistent subdirectory on your cloud drive. The script reads the real project path from
session transcripts to derive a stable cloud subdirectory:
~/git/my-project → <cloud-root>/git/my-project
~/fbsource → <cloud-root>/fbsource (same on any machine)
A Stop hook runs after every response to catch newly created project directories.
The hook exits immediately when nothing needs doing — no overhead in the common case.
Requirements
bashjq- A cloud drive (Google Drive, Dropbox, iCloud, etc.)
Setup
After installing, run /memory-sync-setup in any Claude Code session. Repeat on each
machine, pointing to the same cloud root path.
Projects outside $HOME
If your projects live outside $HOME (e.g. /data/users/$USER/), add the prefix to
the known_prefixes array in sync-memory.sh. The script only links projects whose
real path matches a known prefix.