Website Monitoring for Developers: API, CLI, MCP

Short answer. Developers don't want a click-through dashboard — they want monitoring as code. enterno.io gives you a REST API документацию (v4) to manage monitors from CI/CD, curl-based CLI SNI for health checks, and an MCP server with 16 tools for Claude, Cursor, and Zed. You can create, read, and delete monitors programmatically, embed checks in your pipeline, and ask your assistant "is my API down" right inside the IDE.

Monitoring as code

Adding monitors by hand through a UI doesn't scale: at 50 services your configuration drifts from reality. The monitoring-as-code approach fixes this — monitors are described in your repo and applied via the API on deploy.

If infrastructure is code, its observability should be code too. A monitor created by hand and forgotten is technical debt that surfaces during an incident.

1. REST API

The v4 API covers full CRUD over monitors. Create a monitor in CI after a successful deploy:

curl -X POST https://enterno.io/api/v4/monitors \
  -H "X-API-Key: $ENTERNO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://api.example.com/health",
    "check_type": "http",
    "interval_minutes": 1,
    "expected_code": 200,
    "notify_telegram": true
  }'

The key is passed via the $ENTERNO_API_KEY environment variable — no secrets in code. You also get GET to list, PUT to edit, and DELETE to remove monitors.

2. CLI and pipeline health checks

Before registering a monitor, it's worth probing the endpoint straight from the pipeline and failing the deploy if it doesn't respond:

#!/usr/bin/env bash
set -euo pipefail
code=$(curl -sS -o /dev/null -w "%{http_code}" https://api.example.com/health)
if [ "$code" != "200" ]; then
  echo "Health check failed: $code" >&2
  exit 1
fi
echo "Healthy"

The script returns a non-zero exit code on failure — CI halts the deploy before a broken build reaches production.

3. MCP server

The most interesting part for anyone working with AI assistants. enterno.io ships an MCP server with 16 tools, compatible with Claude, Cursor, and Zed. The config goes into your client's settings:

{
  "mcpServers": {
    "enterno": {
      "url": "https://enterno.io/mcp"
    }
  }
}

Once connected, you can ask in chat "check DNS for example.com" or "what's my API's uptime over the last day" — the assistant calls the right tool and returns the data. A deeper walkthrough is in the MCP server for monitoring.

Interface comparison

Interface	When to use	Auth
REST API	CI/CD, IaC, bulk management	X-API-Key
CLI (curl)	Pipeline health checks, scripts	none for probing
MCP	Working from the IDE with an AI assistant	via MCP client
Web UI	One-off setup, overview	session

Check types

• HTTP — status code, headers, body, redirects, response time.
• SSL/TLS — certificate expiry, chain, protocol, ciphers.
• Ping/port — ICMP reachability and TCP port openness.
• DNS — A/AAAA/MX/NS records matching expected values.

Every interface supports four monitor types: HTTP (code and response time), SSL (certificate expiry), Ping (ICMP host reachability), and DNS (record resolution). For a backend, HTTP + SSL usually suffices; Ping is useful for infrastructure; and a DNS monitor catches domain issues before users do. More on DNS in the DNS records guide.

Alerts for teams

Notifications go to Telegram, Slack, webhook, PagerDuty, and Jira. For a dev team, Slack (working hours) + PagerDuty (on-call overnight) is the sweet spot. A webhook lets you forward the event to your own service and, say, auto-create a ticket.

FAQ

Can I manage monitors without opening the UI?

Yes. The v4 REST API covers the full monitor lifecycle — create, read, update, delete. The UI is optional.

Which IDEs support the MCP server?

Claude (Desktop and Code), Cursor, and Zed — any client that supports the Model Context Protocol over a URL.

Do I need an API key for a curl health check?

No, not if you're just probing a public endpoint. The key is only required to manage enterno.io monitors via the API.

How many monitors are on the free plan?

Up to 10 monitors with check intervals from 1 minute — enough for a small set of services and side projects.

Wire monitoring into your stack

Start with the REST API or MCP: go to monitors. For notification best practices, see this guide; for uptime fundamentals, the general guide.