Move your AI tool configs to a new machine

New laptop, fresh OS install, or a work/personal machine split: instead of recreating every AI tool’s configuration by hand, move your syllago library once and reinstall from it.

Your library is plain files under ~/.syllago/ — there’s no database or hidden state — so migration is a copy plus a reinstall.

1. On the old machine: copy your library

Transfer the ~/.syllago/ directory however you move files between machines:

# Example: straight over SSH
rsync -a ~/.syllago/ new-machine:~/.syllago/

That directory contains your content (content/ — rules, skills, agents, hooks, commands, and loadouts) plus configuration. Registry clones don’t need to travel — you’ll re-add registries on the new machine and they’ll re-clone.

Tip

If you expect to do this more than once, put your library in a personal git registry instead: scaffold one with syllago registry create --new my-library, push it somewhere private, and a new machine becomes registry add + sync + install — no file copying at all. See Team Setup for the same pattern at team scale.

2. On the new machine: install syllago

Follow Installation, then confirm:

syllago version

3. Verify the library arrived

syllago list

You should see your rules, skills, and other items grouped by type, exactly as on the old machine. If the list is empty, the ~/.syllago/ copy didn’t land where syllago expects — check the path.

4. Re-add your registries

Registry clones are local caches, so add each registry again:

syllago registry add https://github.com/your-org/syllago-registry.git
syllago registry sync

syllago registry list on the old machine tells you what you had. A registry outside syllago’s bundled allowlist also needs the --moat or signing flags you added it with. For a GitHub-hosted registry, that means --signing-identity plus --signing-repository-id and --signing-repository-owner-id. See Registry Signing Identity.

5. Reinstall content to your providers

Initialize in your project (syllago init detects which AI tools are installed on the new machine), then push library content back out:

# Preview first
syllago install --all --to claude-code --dry-run

# Install everything to one provider
syllago install --all --to claude-code

# Or fan out to every detected provider
syllago install --all --to-all

If you use loadouts, they traveled with your library — syllago loadout apply <name> --keep recreates a whole configuration in one command.

6. Verify

syllago list
syllago loadout status

Then open one of your AI tools and confirm it sees its rules — for example, Claude Code should show your rules in context from .claude/rules/ or CLAUDE.md.

Related

• Library — what lives in ~/.syllago/content/
• syllago install — bulk install flags
• Team Setup — the registry-based version of this workflow