| name | dockge |
| description | Manage Docker compose stacks via Dockge — a file-based stack manager that stores stacks as compose files under /opt/stacks. Covers its file-on-disk model, the lack of a REST API/MCP (so operations are filesystem- or SSH-oriented), and how to edit stacks safely. Load this when the user manages Docker with Dockge. |
Dockge
Dockge (https://github.com/louislam/dockge) is a lightweight, file-based compose manager. It's popular in
homelabs but, unlike Dockhand/Portainer, exposes no documented REST API and no MCP server — its UI talks
over Socket.IO. Practical management therefore leans on the filesystem (and SSH). Obey
the homelab-core:homelab-safety skill and load
docker-compose-stacks.
The model
- Stacks live as real files:
${DOCKGE_STACKS_DIR} (default /opt/stacks), one dir per stack containing
compose.yaml + .env.
- Because stacks are plain files, they're friendly to Git/backup and to direct editing — but edits and deploys
generally happen via the Dockge UI or on the host filesystem.
- Agent system (v1.4.0+) lets one Dockge manage multiple hosts, each with its own
/opt/stacks.
Working with it
Since there's no API/MCP:
- Read: you can read
compose.yaml/.env under /opt/stacks/<name> and docker compose ps/logs to
inspect — read-only, fine.
- Change a stack: editing the compose file on disk and (re)deploying is a host filesystem operation →
treat as significant. If it requires SSH to the host, clear the SSH-confirmation gate
() and follow the SSH execution discipline ():
confirm the host, , validate with , then
deploy via the Dockge UI or in that stack dir. Keep the UI and disk in sync.
