Deployment

Deployment Options

portainer-agent exposes its MCP server (console script portainer-mcp) four ways. Pick the row that matches where the server runs relative to your MCP client, then copy the matching mcp_config.json below. Replace the <your-…> placeholders with the values from the Configuration / Environment Variables section.

#	Option	Transport	Where it runs	mcp_config.json key
1	stdio	stdio	client launches a subprocess	command
2	Streamable-HTTP (local)	streamable-http	a local network port	command or url
3	Local container / uv	stdio or streamable-http	Docker / Podman / uv on this host	command or url
4	Remote URL	streamable-http	a remote host behind Caddy	url

1. stdio (local subprocess)

The client launches the server over stdio via uvx — best for local IDEs (Cursor, Claude Desktop, VS Code):

{
  "mcpServers": {
    "portainer-mcp": {
      "command": "uvx",
      "args": ["--from", "portainer-agent", "portainer-mcp"],
      "env": {
        "PORTAINER_URL": "<your-portainer_url>",
        "PORTAINER_ENDPOINT": "<your-portainer_endpoint>",
        "PORTAINER_USERNAME": "<your-portainer_username>"
      }
    }
  }
}

2. Streamable-HTTP (local process)

Run the server as a long-lived HTTP process:

uvx --from portainer-agent portainer-mcp --transport streamable-http --host 0.0.0.0 --port 8000
curl -s http://localhost:8000/health        # {"status":"OK"}

Then either let the client launch it:

{
  "mcpServers": {
    "portainer-mcp": {
      "command": "uvx",
      "args": ["--from", "portainer-agent", "portainer-mcp", "--transport", "streamable-http", "--port", "8000"],
      "env": {
        "TRANSPORT": "streamable-http",
        "HOST": "0.0.0.0",
        "PORT": "8000",
        "PORTAINER_URL": "<your-portainer_url>",
        "PORTAINER_ENDPOINT": "<your-portainer_endpoint>",
        "PORTAINER_USERNAME": "<your-portainer_username>"
      }
    }
  }
}

…or connect to the already-running process by URL:

{
  "mcpServers": {
    "portainer-mcp": { "url": "http://localhost:8000/mcp" }
  }
}

3. Local container / uv

(a) Launch a container directly from mcp_config.json (stdio over the container — no ports to manage). Swap docker for podman for a daemonless runtime:

{
  "mcpServers": {
    "portainer-mcp": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "TRANSPORT=stdio",
        "-e", "PORTAINER_URL=<your-portainer_url>",
        "-e", "PORTAINER_ENDPOINT=<your-portainer_endpoint>",
        "-e", "PORTAINER_USERNAME=<your-portainer_username>",
        "knucklessg1/portainer-agent:latest"
      ]
    }
  }
}

(b) Run a local streamable-http container, then connect by URL:

docker run -d --name portainer-mcp -p 8000:8000 \
  -e TRANSPORT=streamable-http \
  -e PORT=8000 \
  -e PORTAINER_URL="<your-portainer_url>" \
  -e PORTAINER_ENDPOINT="<your-portainer_endpoint>" \
  -e PORTAINER_USERNAME="<your-portainer_username>" \
  knucklessg1/portainer-agent:latest
# or, from a clone of this repo:
docker compose -f docker/mcp.compose.yml up -d

{
  "mcpServers": {
    "portainer-mcp": { "url": "http://localhost:8000/mcp" }
  }
}

(c) From a local checkout with uv:

uv run portainer-mcp --transport streamable-http --port 8000

4. Remote URL (deployed behind Caddy)

When the server is deployed remotely (e.g. as a Docker service) and published through Caddy on the internal *.arpa zone, connect with the "url" key — no local process or image required:

{
  "mcpServers": {
    "portainer-mcp": { "url": "http://portainer-mcp.arpa/mcp" }
  }
}

Caddy reverse-proxies http://portainer-mcp.arpa to the container's :8000 streamable-http listener; http://portainer-mcp.arpa/health returns {"status":"OK"} when the service is live.

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

portainer-agent ships two console scripts: an MCP server (portainer-mcp) that exposes the typed Portainer tool surface, and an A2A agent server (portainer-agent) that routes natural-language requests to those tools through a graph orchestrator.

Run the MCP server

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

stdio (default)streamable-httpsse

portainer-mcp

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

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

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

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

Health check (HTTP transports):

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

Configuration (environment)

portainer-agent is configured entirely from the environment. The required set:

Var	Default	Meaning
PORTAINER_URL	http://localhost:9000	Portainer instance URL
PORTAINER_TOKEN	(empty)	API access token (X-API-Key)
PORTAINER_USERNAME	admin	Username for username/password auth
PORTAINER_PASSWORD	(empty)	Password for username/password auth
PORTAINER_SSL_VERIFY	True	Verify TLS (set False for self-signed homelab)
HOST	0.0.0.0	Bind address (HTTP transports)
PORT	8000	Bind port (HTTP transports)
TRANSPORT	stdio	stdio, streamable-http, or sse

Each management domain is gated by a *TOOL toggle — AUTHTOOL, ENVIRONMENTTOOL, DOCKERTOOL, STACKTOOL, KUBERNETESTOOL, EDGETOOL, TEMPLATETOOL, USERTOOL, REGISTRYTOOL, SYSTEMTOOL (all True by default). The full set, including telemetry (OTEL) and access-governance (Eunomia) options, is documented in .env.example. Copy it to .env and fill in 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:
  portainer-agent-mcp:
    image: knucklessg1/portainer-agent:latest
    container_name: portainer-agent-mcp
    hostname: portainer-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 PORTAINER_* values
docker compose -f docker/mcp.compose.yml up -d
docker compose -f docker/mcp.compose.yml logs -f

A2A agent server

portainer-agent also ships an A2A agent server (console script portainer-agent). It connects to the MCP server over HTTP and exposes a natural-language interface with an optional web UI.

export PORTAINER_URL=http://your-portainer:9000
export PORTAINER_TOKEN=your_api_token
portainer-agent --provider openai --model-id gpt-4o --api-key sk-...

The repo ships docker/agent.compose.yml, which runs the MCP server and the agent together. The agent listens on :9004 and reaches the MCP server by container name via MCP_URL:

services:
  portainer-agent-mcp:
    image: knucklessg1/portainer-agent:latest
    container_name: portainer-agent-mcp
    hostname: portainer-agent-mcp
    restart: always
    env_file:
      - ../.env
    environment:
      - PYTHONUNBUFFERED=1
      - HOST=0.0.0.0
      - PORT=8000
      - TRANSPORT=streamable-http
    ports:
      - "8000:8000"

  portainer-agent-agent:
    image: knucklessg1/portainer-agent:latest
    container_name: portainer-agent-agent
    hostname: portainer-agent-agent
    restart: always
    depends_on:
      - portainer-agent-mcp
    env_file:
      - ../.env
    command: ["portainer-agent"]
    environment:
      - PYTHONUNBUFFERED=1
      - HOST=0.0.0.0
      - PORT=9004
      - MCP_URL=http://portainer-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

Behind a Caddy reverse proxy

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

# Internal (self-signed) — homelab .arpa zone
portainer-agent.arpa {
    tls internal
    reverse_proxy portainer-agent-mcp:8000
}

# Public — automatic Let's Encrypt
portainer-agent.example.com {
    reverse_proxy portainer-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.arpa:5380/api/zones/records/add" \
  --data-urlencode "token=$TECHNITIUM_DNS_TOKEN" \
  --data-urlencode "domain=portainer-agent.arpa" \
  --data-urlencode "zone=arpa" \
  --data-urlencode "type=A" \
  --data-urlencode "ipAddress=10.0.0.10" \
  --data-urlencode "ttl=3600"

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

Register with an MCP client

Add to your client's mcp_config.json (multiplexer nickname pt):

{
  "mcpServers": {
    "portainer-agent": {
      "command": "uv",
      "args": ["run", "portainer-mcp"],
      "env": {
        "PORTAINER_URL": "http://your-portainer:9000",
        "PORTAINER_TOKEN": "your_api_token",
        "PORTAINER_SSL_VERIFY": "True"
      }
    }
  }
}

For a remote HTTP server, point the client at http://portainer-agent.arpa/mcp instead.