Getting started
Create the first folder and hub, then join an empty destination on another machine.
The interactive wizard is the shortest path. It validates the same constraints as automation flags and can install the background service when the initial synchronization succeeds.
Before you start
- Choose one parent folder whose direct children are Git repositories.
- Choose a local hub path or an SSH hub path on a trusted machine.
- Keep hub and peer state outside the synchronized root.
- Install the same Code Folder Sync version on every participating machine.
Create the folder
codefoldersync setupChoose create, provide the root and hub, give the folder and peer readable names, and allow the filesystem probes to complete. The result prints a folder ID.
Automation can use the same validation path:
codefoldersync setup \
--mode create \
--root /absolute/path/to/code \
--hub ssh://user@hub-host/absolute/path/to/hub \
--name my-code \
--peer laptopJoin another machine
The destination must be empty. This makes an initial join distinguishable from an untracked local tree and gives interrupted materialization a safe recovery boundary.
codefoldersync setup \
--mode join \
--folder-id <folder-id> \
--root /absolute/path/to/empty/code \
--hub ssh://user@hub-host/absolute/path/to/hub \
--peer desktopVerify and start
codefoldersync doctor
codefoldersync status
codefoldersync verify --full
codefoldersync service install
codefoldersync service startverify --full deliberately rehashes the complete tree. It is the authoritative expensive check after a first join, restore, conflict resolution, or incident.