From bridgey-deploy
Sync agent files between local and remote server. Use when the user says "sync", "push to remote", "pull from remote", "sync agent", or runs /sync.
How this skill is triggered — by the user, by Claude, or both
Slash command
/bridgey-deploy:syncThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Bidirectional rsync between local agent and remote deployment. Connection details are read from `bridgey-deploy.config.json` in the agent's root directory.
Bidirectional rsync between local agent and remote deployment. Connection details are read from bridgey-deploy.config.json in the agent's root directory.
Read bridgey-deploy.config.json to get:
tailscale_host — the Tailscale hostname of the remote serverremote_path — the path on the remote server (e.g., /opt/bridgey/personas/{name})container_name — the Docker container nameIf bridgey-deploy.config.json doesn't exist, ask the user for the Tailscale hostname and remote path.
/sync push — local -> remote (config, skills, CLAUDE.md changes)/sync pull — remote -> local (memory, outputs, learnings)/sync — bidirectional (push then pull)Push local config/skills/rules to the remote server. Excludes runtime data.
rsync -avz --delete \
--exclude='.git/' \
--exclude='.mcp.json' \
--exclude='*.log' \
--exclude='*.db' \
--exclude='*.db-journal' \
--exclude='user/memory/' \
~/.personas/{name}/ {tailscale_host}:{remote_path}/
After pushing, restart the container to pick up changes:
ssh {tailscale_host} "cd $(dirname {remote_path}) && docker compose restart {container_name}"
Pull memory and outputs from the remote server. Only syncs user/memory/.
rsync -avz \
{tailscale_host}:{remote_path}/user/memory/ \
~/.personas/{name}/user/memory/
Run push then pull in sequence.
.mcp.json is NEVER synced (contains API keys, differs per environment).git/ is never synceduser/memory/ (remote writes there, local pulls it)npx claudepluginhub kickinrad/bridgey --plugin bridgey-deployImports an OpenClaw agent workspace (SOUL.md, IDENTITY.md, MEMORY.md, etc.) into the current project. Useful when migrating or referencing an agent's identity and memory.
Manually syncs Claude brain state across machines: checks initialization, pushes local changes, pulls and auto-merges remote updates, shows summary, suggests conflict resolution.