create-port-monitor

Create a PORT monitor

Preflight — read first. If you cannot see any uptimerobot:* MCP tools in your tool list, invoke the uptimerobot:setup skill before doing anything else. Do not tell the user the MCP is misconfigured — setup's Step 0 detects the common case (server connected, tools loaded after session start) and resolves it without re-keying.

Use when the user wants to check that a specific TCP port is open on a host — databases, SSH, SMTP, custom services, etc.

Tool: create-monitor Required params: friendlyName, type: "PORT", url (host or IP, no scheme), port (1–65535).

Minimal call

{
  "friendlyName": "Postgres primary",
  "type": "PORT",
  "url": "db.example.com",
  "port": 5432
}

Common optional params

• timeout — how long to wait for the TCP handshake (1–60s).
• interval, responseTimeThreshold, assignedAlertContacts, tagNames, maintenanceWindowsIds.
• config.ipVersion — IPv4 or IPv6.

Examples

SMTP over IPv4:

{
  "friendlyName": "Mail relay",
  "type": "PORT",
  "url": "smtp.example.com",
  "port": 25,
  "timeout": 15,
  "interval": 300
}

SSH with alert contacts:

{
  "friendlyName": "Bastion SSH",
  "type": "PORT",
  "url": "bastion.example.com",
  "port": 22,
  "assignedAlertContacts": [
    { "alertContactId": "10", "threshold": 0, "recurrence": 0 }
  ],
  "tagNames": ["infra", "security"]
}

Common mistakes

• Passing https://host:443 as url. Split into url: "host" and port: 443.
• Passing port as a string. It must be an integer 1–65535.
• Passing port outside the valid range (e.g. 0 or 70000). The server rejects with -29001.
• Using PORT for an HTTP health check where you actually want status code / body — prefer HTTP or KEYWORD.

After creation

Call get-monitor-details to confirm host + port were stored.

Related

• create-udp-monitor — for UDP services (DNS, NTP, QUIC, game servers).
• create-ping-monitor — for reachability without a specific port.