Show HN: An uptime monitor with an MCP server, create monitors by asking Claude

Monitive launched an API-first uptime monitoring service with an MCP server that lets developers and AI agents create monitors via curl, Claude prompts, or GitHub Actions. The platform offers 50 free monitors, evidence-rich incident data, and multiple alert channels including webhooks, browser push, RSS, and MCP pull.

By Monitive · 16 years in monitoring · 50 monitors free · GitHub login · no email · no card API-first uptime monitoring for deploy pipelines, developers, and AI agents. Create monitors with a curl call, an MCP prompt, or a GitHub Action. claude mcp add --transport http uptimemonitoring \ https://api.uptimemonitoring.com/mcp Have an API key? Add --header "Authorization: Bearer umk live ..." to authorize via key. Use the surface you already work in. Same backend. Same monitor state. Different ways to drive it. Create and manage monitors from scripts, apps, or CI. curl -X POST https://api.uptimemonitoring.com/api/v1/monitors \ -H "Authorization: Bearer $UPTIMEMONITORING API KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "api-prod", "url": "https://api.example.com/healthz", "type": "http" }' Built for your workflow Pick the surface that fits how you ship. Fail deploys when your site is down Use a GitHub Action to assert your deploy is reachable from 22 probe locations before traffic switches over. Let your Claude agent monitor its own deployments Add the UptimeMonitoring MCP server to Claude Code so your agent can create monitors, check status, and react to failures. Uptime monitoring from inside Cursor Configure UptimeMonitoring as an MCP server in Cursor and manage monitors from the agent panel. Monitor any API with 50 free checks Create monitors via the REST API with a single POST request — no dashboard required. What happens when something breaks We do not flip red on the first blip. Create You create a monitor through the API, MCP, or GitHub Action. Test immediately We run an immediate test check so typos and unreachable targets fail fast. Check on schedule Checks are spread across the minute to keep load stable. Re-check failures Infrastructure-class failures are confirmed from two additional regions before we flip state. Deliver evidence We store the incident with evidence and send it where you want it: webhook, browser push, RSS, or MCP pull. Evidence, not just red dots See what happened before the incident: DNS, TLS, timeout, 5xx, region, and timing breakdowns. - Last 20 checks included - Timing breakdowns per check - Region-aware evidence - Designed for debugging, not dashboards { "id": 194, "type": "down", "monitor id": 1287, "started at": "2026-04-18T05:38:52Z", "resolved at": "2026-04-18T05:45:44Z", "evidence": { "region": "US-W", "timestamp": "2026-04-18T05:37:41Z", "status code": 200, "dns ms": 6.8, "tls ms": 298, "connect ms": 149, "ttfb ms": 203 }, { "region": "US-W", "timestamp": "2026-04-18T05:38:42Z", "error class": "dns" } } Webhooks for systems. Push for humans. MCP for agents. No inbox spam. No deliverability headaches. You choose the channel. Webhooks For Slack, Twilio, Discord, ntfy, GitHub Actions, or your own router. PUT /api/v1/monitors/1287/webhook { "url": "https://hooks.co/inc" } Browser push One-click opt-in from the dashboard. Fastest human fallback. Enable notifications Chrome on macOS · Active RSS / incident feed Subscribe from Slack, readers, or automation tools. https://api.uptimemonitoring.com /feed/{token}/rss MCP pull Ask your agent for monitor state and recent incidents directly. Tool: list incidents Filter: last 24h, status: open Result: 1 incident, myapp-healthz Browser push and email get muted by Do Not Disturb — exactly when an outage matters. Wire a webhook to Pushover bypasses DND on iOS and Android or ntfy for phone alerts the moment something breaks. Reliable alerts guide → /guides/reliable-alerts Validate the things that actually break HTTP / HTTPS Status codes, body match, timing evidence Timing evidence DNS, connect, TLS, TTFB, download Copy a working example Start from a real recipe, not a blank page. Free plan Enough for real projects. Simple enough to remember. - Up to 50 monitors - Up to 100 during extended free tier early users - 60-second minimum interval - 22 global probe locations - 1 webhook per monitor - 10 API keys per account - 30-day incident retention - Full MCP access - Browser push + RSS incident feed - No email required - No credit card required Paid plans will add 30-second intervals, longer retention, and response-time threshold alerts. Built by Monitive UptimeMonitoring.com is the API-first, headless sibling to Monitive Pro. Monitive has been operating monitoring infrastructure for 16 years. UptimeMonitoring.com takes that operational experience and applies it to a different job: deploy gates, API workflows, AI-agent tooling, and developer-first monitoring without email or dashboard sprawl. - 16 years in monitoring - 22 probe locations - Cross-region confirmation on failures - GitHub OAuth + API keys - Minimal data collection: GitHub identity only API / MCP / GitHub Action │ ▼ Go dispatcher + evidence │ ▼ 22-probe network │ ▼ webhooks / push / RSS / incidents FAQ Is it really free? Yes. Free plan: up to 50 monitors. Early users can go up to 100 during the extended free tier while paid plans are being finalized. Do I need an email address? No. Signup uses GitHub OAuth. Email is optional only if you want future pricing updates. How do alerts work without email? Webhooks, browser push, RSS feeds, and MCP queries. How do you avoid false positives? Infrastructure-class failures are re-checked across regions before state changes. Can I monitor private IPs or internal services? No. Private and reserved targets are blocked.