Deployment¶

Deployment Options¶

uptime-kuma-agent supports local stdio, a loopback-only development listener, a least-privilege stdio container, and a remote authenticated HTTPS boundary. Provider endpoint, credential, selector, identity, and trust material are supplied at runtime through AgentConfig; none is stored in this repository.

Installed stdio process¶

{
  "mcpServers": {
    "uptime": {
      "command": "uptime-mcp",
      "args": [],
      "env": {"MCP_TOOL_MODE": "intent"}
    }
  }
}

Loopback development listener¶

uptime-mcp --transport streamable-http --host 127.0.0.1 --port 8000

Do not expose this listener beyond loopback. Network deployments require direct TLS or an explicitly trusted TLS-terminating ingress, configured authentication, exact MCP_ALLOWED_HOSTS, and an exact trusted-proxy CIDR policy.

Least-privilege local container¶

docker run -i --rm \
  --read-only \
  --cap-drop=ALL \
  --security-opt=no-new-privileges \
  --pids-limit=256 \
  --tmpfs /tmp:rw,noexec,nosuid,nodev,size=64m \
  -e TRANSPORT=stdio \
  registry.example.invalid/uptime-kuma-agent@sha256:<digest> uptime-mcp

The operator projects the selected AgentConfig profile into the process at runtime; the image remains immutable and contains no environment connection profile.

Remote authenticated HTTPS endpoint¶

{
  "mcpServers": {
    "uptime": {"url": "https://service.example.invalid/mcp"}
  }
}

Store the real remote URL, outbound identity reference, and TLS-profile reference in AgentConfig, not in MCP client JSON or documentation.

This page covers running uptime-kuma-agent as a long-lived service: the transports, a Docker Compose stack, the optional A2A agent server, putting it behind a Caddy reverse proxy, and giving it a DNS name with Technitium. To provision the Uptime Kuma instance it connects to, see Backing Platform.

uptime-kuma-agent ships two console scripts: the MCP server (uptime-mcp) and a Pydantic AI agent server (uptime-agent). The MCP server is a typed, deterministic tool surface a policy router / agent calls; the agent server adds a conversational graph agent and web UI on top of it.

Run the MCP server¶

The transport is selected with --transport (or the TRANSPORT env var):

stdio (default)streamable-httpsse

uptime-mcp

For IDE / desktop MCP clients that launch the server as a subprocess.

uptime-mcp --transport streamable-http --host 0.0.0.0 --port 8000

A network server with a /health endpoint and /mcp route.

uptime-mcp --transport sse --host 0.0.0.0 --port 8000

Health check (HTTP transports):

curl -s http://localhost:8000/health        # {"status":"OK"}

Configuration (environment)¶

uptime-kuma-agent is configured entirely from the environment. The required connection set:

Var	Default	Meaning
`UPTIME_KUMA_URL`	`http://localhost:3001`	Base URL of the Uptime Kuma instance
`AUTH_TYPE`	(unset)	`password` or `token`
`UPTIME_KUMA_USERNAME`	(unset)	Login user (when `AUTH_TYPE=password`)
`UPTIME_KUMA_PASSWORD`	(unset)	Login password (when `AUTH_TYPE=password`)
`UPTIME_KUMA_TOKEN`	(unset)	Access token (when `AUTH_TYPE=token`)
`HOST`	`0.0.0.0`	Bind address for HTTP transports
`PORT`	`8000`	Bind port for HTTP transports
`TRANSPORT`	`stdio`	`stdio`, `streamable-http`, or `sse`
`MONITORSTOOL`	`True`	Register the `uptime_kuma_monitors` tool
`STATUSTOOL`	`True`	Register the `uptime_kuma_status` tool
`ENABLE_OTEL`	`True`	Export traces via OpenTelemetry
`EUNOMIA_TYPE`	`none`	Policy enforcement: `none`, `embedded`, `remote`

The complete set, including the OpenTelemetry and Eunomia options, is documented in .env.example. Copy it to .env and populate only what you use.

Docker Compose¶

The repo ships docker/mcp.compose.yml. It reads a sibling .env and publishes the HTTP server on :8000:

services:
  uptime-kuma-agent-mcp:
    image: example/uptime-kuma-agent@sha256:<digest>
    container_name: uptime-kuma-agent-mcp
    hostname: uptime-kuma-agent-mcp
    restart: always
    env_file:
      - ../.env
    environment:
      - PYTHONUNBUFFERED=1
      - HOST=0.0.0.0
      - PORT=8000
      - TRANSPORT=streamable-http
    ports:
      - "8000:8000"
    healthcheck:
      test: ["CMD", "python3", "-c", "import urllib.request; urllib.request.urlopen('http://localhost:8000/health')"]
      interval: 30s
      timeout: 10s
      retries: 3

cp .env.example .env          # then edit UPTIME_KUMA_* values
docker compose -f docker/mcp.compose.yml up -d
docker compose -f docker/mcp.compose.yml logs -f

Agent server¶

For the conversational A2A experience, the repo ships docker/agent.compose.yml, which deploys the MCP server and the uptime-agent graph agent together. The agent serves its web UI on :9004 and is wired to the MCP server through MCP_URL:

services:
  uptime-kuma-agent-agent:
    image: example/uptime-kuma-agent@sha256:<digest>
    container_name: uptime-kuma-agent-agent
    hostname: uptime-kuma-agent-agent
    restart: always
    depends_on:
      - uptime-kuma-agent-mcp
    env_file:
      - ../.env
    command: [ "uptime-agent" ]
    environment:
      - PYTHONUNBUFFERED=1
      - HOST=0.0.0.0
      - PORT=9004
      - MCP_URL=http://uptime-kuma-agent-mcp:8000/mcp
      - PROVIDER=${PROVIDER:-openai}
      - MODEL_ID=${MODEL_ID:-gpt-4o}
      - ENABLE_WEB_UI=True
    ports:
      - "9004:9004"

docker compose -f docker/agent.compose.yml up -d

Run the agent directly from the command line for interactive use:

uptime-agent --provider openai --model-id gpt-4o

Behind a Caddy reverse proxy¶

Expose the HTTP server on a hostname with automatic TLS. Add to your Caddyfile:

# Internal (self-signed) — homelab .example.invalid zone
uptime-kuma-agent.example.invalid {
    tls internal
    reverse_proxy uptime-kuma-agent-mcp:8000
}

# Public — automatic Let's Encrypt
uptime-kuma-agent.example.com {
    reverse_proxy uptime-kuma-agent-mcp:8000
}

Reload Caddy:

docker compose -f services/caddy/compose.yml exec caddy caddy reload --config /etc/caddy/Caddyfile

DNS with Technitium¶

Point the hostname at the host running Caddy. Via the Technitium API:

curl -s "http://technitium.example.invalid:5380/api/zones/records/add" \
  --data-urlencode "token=$TECHNITIUM_DNS_TOKEN" \
  --data-urlencode "domain=uptime-kuma-agent.example.invalid" \
  --data-urlencode "zone=arpa" \
  --data-urlencode "type=A" \
  --data-urlencode "ipAddress=192.0.2.10" \
  --data-urlencode "ttl=3600"

…or add an A record uptime-kuma-agent.example.invalid → <caddy-host-ip> in the Technitium web console (http://technitium.example.invalid:5380). The ecosystem technitium-dns-mcp automates this as a tool.

Register with an MCP client¶

Add to your client's mcp_config.json:

{
  "mcpServers": {
    "uptime-kuma-agent": {
      "command": "uv",
      "args": ["run", "uptime-mcp"],
      "env": {
        "UPTIME_KUMA_URL": "http://localhost:3001",
        "AUTH_TYPE": "password",
        "UPTIME_KUMA_USERNAME": "admin",
        "UPTIME_KUMA_PASSWORD": "your_password_here"
      }
    }
  }
}

For a remote HTTP server, point the client at http://uptime-kuma-agent.example.invalid/mcp instead.