Agentic control for self-hosted apps

Run your homelab with an agent you control.

Stackarr gives trusted agents a typed, safety-controlled way to operate apps, containers, downloads, backups, and infrastructure. You keep the dashboard and the final say.

Start with an agentSee safety profilesInstall Stackarr

Start with approvals. Expand authority only when the workflow earns your trust.

Stackarr control planemanage
You

Check the stack, fix what is safe, and ask before interrupting playback.

Agent plan3 actions
  1. 01Inspect enabled servicesStackarr + native app APIsread
  2. 02Repair stalled downloadAllowed by manage profilerun
  3. 03Restart PlexPlayback may be interruptedask
Tools selected from installed apps
AgregarrImmichPlexPulsarrRadarrSonarrTracearr

Bring your own agent

Your tools, one shared operating model

Every client reaches the same actions and safety policy. Stackarr exposes only the apps you installed, so an agent gets a focused toolset with clear consequences.

CodexClaudeHermesOpenClawLM StudioAny stdio MCP client
  1. 01

    Connect once

    Add Stackarr as a local MCP server, including native MCP connections for Hermes and OpenClaw.

    Open agent setup →
  2. 02

    Choose authority

    Use observe, manage, admin, or unrestricted. The launch profile, not the agent, sets the boundary.

    Compare MCP profiles →
  3. 03

    Speak normally

    Ask for an install, a health check, a media request, a repair, or a full stack migration.

    Connect Hermes or OpenClaw →

A complete manual surface

See the stack clearly. Take over at any time.

Setup, status, activity, recovery, and direct app access live in the same place. Human and agent actions share one source of truth.

Stackarr dashboard showing active work, running containers, and live CPU and memory history
Stackarr Apps screen

Apps

Open the apps you installed, pin the ones you use most, and keep each connection and behavior setting close to its app.

Stackarr UI settings screen

UI settings

Tune the experience to your setup with theme, refresh, service-link, and onboarding preferences that stay with the app.

Private by default

Install at home. Connect from the tools you use.

1. Start Stackarr

Download the Compose file and run the control plane on any Docker-compatible host.

curl -fsSL https://stackarr.app/docker-compose.yml \
  -o docker-compose.yml
docker compose --profile stackarr up -d app

2. Connect your assistant

Start with approval-backed management, use admin for setup, or deliberately grant full autonomy.

codex mcp add stackarr -- \
  docker exec -i -e STACKARR_MCP_PROFILE=manage \
  app /app/bin/stackarr mcp serve

See the MCP connection guides for Claude, Hermes, OpenClaw, and LM Studio.

One control plane

Operate the same stack from chat, the dashboard, the API, or the CLI without splitting state between tools.

Actions built for each app

Agents use typed Stackarr and app APIs instead of improvising shell commands against your containers.

Boundaries you choose

Start read-only, allow routine management, unlock setup, or deliberately grant unrestricted control.

Evidence for every action

Review plans, approvals, results, and failures in one activity trail whenever you want to take over.