# KiteStacks Cloud Migration Backup

Created for migrating the current KiteStacks host to Hetzner Cloud.

## Contents

- `archives/docker-bind-data.tar.gz` - `/home/kenpat/docker` bind-mounted service folders, including compose files, `.env` files, and bind-mounted app data.
- `archives/syncthing-shared.tar.gz` - `/home/kenpat/SyncthingShared`.
- `archives/kitestacks-scripts.tar.gz` - local KiteStacks automation/script folders.
- `archives/host-etc-subset.tar.gz` - selected host `/etc` configuration needed for migration context.
- `volume-exports/*.tar.gz` - Docker named volume exports.
- `inventory/*` - Docker, network, disk, config, and host inventory.
- `restore/RESTORE.md` - first-pass restore procedure.
- `SHA256SUMS` - checksums for all files in this backup.

## Restoring Claude's Context on a New Machine

This repo includes a snapshot of Claude Code's persistent memory in `claude-memory/`
(taken from `~/.claude/projects/-home-kenpat/memory/` on the original host). It
captures the homelab setup, this migration plan/status, and your working
preferences, so a fresh Claude session on the new machine doesn't start blank.

To restore it:

1. Clone this repo on the new machine.
2. Copy the contents of `claude-memory/` into Claude Code's memory directory for
   your new home path, e.g.:
   ```sh
   mkdir -p ~/.claude/projects/-home-<youruser>/memory
   cp claude-memory/*.md ~/.claude/projects/-home-<youruser>/memory/
   ```
3. Start a Claude Code session in that project. It will automatically load
   `MEMORY.md` and the linked memory files, and can pick up the migration
   (and any other tracked work) where it left off.
4. Update `claude-memory/project-cloud-migration.md` as migration phases
   progress, and re-copy it back into this repo so future restores stay current.

## Important

This repository contains secrets: `.env` files, tunnel tokens, app database data, and service credentials. Keep it private.

The Docker bind-data archive was created from a live host. For databases with strict consistency requirements, prefer restoring service-native dumps when available, or stop services before taking a final cutover backup.