# DockTail Cloud > Hosted monitoring for DockTail. Every DockTail host and service in one view, with incidents that say which layer broke (the container, its Tailscale exposure, or the host) and carry the evidence: the Docker event and the log tail from the moment of failure. Your AI agent can investigate those incidents through an MCP server and a REST API. Web page: https://docktail.org/cloud/ · Dashboard: https://cloud.docktail.org/login · Docs: https://docktail.org/docs/cloud/ ## What it does - **Monitors from your labels.** Every container DockTail publishes through its `docktail.*` labels is monitored automatically. Start a labeled container and it is watched; remove it and its monitor goes with it. - **Two vantages, one answer.** The agent checks each service locally on its Docker IP, and Cloud reads the Tailscale control plane to see whether the service is advertised and approved. Together they tell a container problem from an exposure problem. - **Incidents with evidence.** Exit code, OOM kill, restart loop or health change, plus a bounded, redacted log tail captured when the failure fired. - **Alerts and recoveries** to Slack, Telegram, ntfy, webhooks or email. Hosts that stop reporting are detected in Cloud, so a dead box still alerts. - **The rest of the box.** Host vitals (CPU, memory, disk, load) and a read-only inventory of the host's other containers. - **AI agents.** A remote MCP server and a REST API let Claude Code, Cursor, Codex or a script read hosts, services, incidents and logs and find the root cause. See https://docktail.org/cloud/agents.md. ## Getting started A person has to do the first three steps; an AI agent cannot create an account or start a subscription. 1. Sign up at https://cloud.docktail.org/login. 2. Create a workspace. 3. Choose a plan. Homelab and Team start with a 30-day free trial that needs a card and turns into the paid plan unless you cancel. A workspace connects no host until it has a plan or a running trial. 4. Copy the workspace key (`dtc_...`) and set it as `DOCKTAIL_CLOUD_KEY` on the DockTail container, then recreate it (`docker compose up -d`). Full setup, including prerequisites and troubleshooting: https://docktail.org/docs/cloud/#how-to-enable ## Plans Priced by monitored hosts; services are not metered. Every plan includes API and MCP access. - **Homelab** ($5/month or $30/year, 30-day free trial): Monitor 2 hosts, 30-day retention, 1 alert destination, Single user, API & MCP access. - **Team** ($29/month or $299/year, 30-day free trial): Monitor 10 hosts, 30-day retention, 5 alert destinations, Team members & roles, API & MCP access. - **Business** ($99/month or $999/year): Up to 100 hosts (30 included), 365-day retention, 15 alert destinations, Team members & roles, Priority email support, Advanced alert routing, API & MCP access. Current prices: https://docktail.org/cloud/#pricing ## Security The DockTail agent is open source (AGPL). It sends operational metadata over one outbound-only, encrypted WebSocket. The protocol has no exec, deploy or shell messages, and Tailscale credentials never leave the host. Details: https://docktail.org/security/ ## FAQ **How does the trial work?** Homelab and Team open with a 30-day trial of the full plan: health checks, incidents, alerts, and API & MCP access. The trial starts at checkout with a card; cancel before it ends and you are not charged. Business starts at its regular price. Every plan includes API & MCP access. **What counts as a host?** One Docker engine running one DockTail agent. Services are not metered: a host reports every service it publishes. Hosts beyond your plan's cap stay visible in the dashboard but are not monitored until you upgrade or remove one; nothing is rejected. **Can I self-host the control plane?** No. The agent is open source under the AGPL; the control plane is a hosted, closed service. Alerting has to live outside the blast radius of the hosts it watches, and a monitor on the box you monitor dies with the box. **What does the agent send, and can Cloud reach into my hosts?** Operational metadata over one outbound-only WebSocket: service snapshots, Docker failure events, local check results, host vitals, Tailscale control-plane state, and bounded, redacted log tails on incidents. The protocol has no exec, deploy, or shell message types, and Tailscale credentials never leave the host. The full list is at docktail.org/security. **Can my AI agent use it?** Yes. DockTail Cloud has a REST API and a remote MCP server on every plan, so Claude Code, Cursor, Codex or a script can list hosts and services, read incidents and their log tails, and, with a read_write key, acknowledge or resolve incidents and open maintenance windows. Neither can run anything on a host. Setup is at docktail.org/cloud/agents. **How is this different from Uptime Kuma?** Uptime Kuma tells you a service is down. DockTail Cloud tells you which layer broke: the container, its exposure on the tailnet, or the host. Its monitors are derived from the docktail.* labels you already wrote, and detection runs outside your network so it keeps working after a host is gone. Full comparison at docktail.org/cloud/vs-uptime-kuma.