TellCopilotClaude to watch your sites and apps.
AI-native monitoring. HTTP uptime, crons, SSL, content checks. Add monitors and triage incidents without leaving your AI assistant.
Essentials
Every monitor you'd duct-tape together. Already here.
Every check on the side of your service, every clock you don't want to forget, every certificate before it bites. One green dot when it's all fine.
Real check, not a ping
HTTP uptime checks
Miss a beat, hear it first
Cron heartbeats
Never expires on you
Domain expiry
Trusted before it bites
SSL certificates expiry
And the rest of the stack
And much, much more.
What's ready today, and what's planned next.
- Public status pages A clean public page in a click. Tells the truth so your inbox doesn't have to.
- Email subscribers Double opt-in, one-click unsubscribe. They hear about incidents before customers ask.
- JSON + RSS feeds Stable status.json and feed.rss for widgets, embeds, and your own dashboards.
- Maintenance windows Schedule the outage. Notdown mutes the right monitors and posts the window for subscribers.
- Mute Silence one monitor for 1h, 4h, 24h, or until you unmute. Checks still record.
- Content assertions Treats a 200 with the word "Database error" the way it deserves. Contains, regex, json-path.
- Email, Telegram, Slack, Discord channels Email, Telegram, Slack, Discord. Pick one or wire up all of them per monitor.
- REST API Sanctum-token REST API at /api/v1. Every monitor verb, every channel, fully scriptable.
- MCP server Plug Notdown into Claude Code, Cursor, Claude Desktop. 36 tools, same auth.
- Status page directory Public directory of status pages, tagged by stack. Search "Laravel" or "Next.js", find every matching page, subscribe in one click.
- Start + finish jobs Long-running jobs ping start and finish. Catches the one that began but never came back.
- TCP + ICMP TCP port reachability and ICMP ping. For the parts of your stack beneath HTTP.
- DNS + DMARC Watch DNS records change underneath you. Alert on missing or weak DMARC policy.
- Blacklist + Safe Browsing Daily check against 15 RBLs and Google Safe Browsing. Know before email starts bouncing.
- Visual regression Screenshot diff against a baseline. Catches the silent deploy that broke the page.
- Multi-region probes Confirm a fail from EU and US before paging you. Eliminates one-off network noise.
- Failure screenshots First failed check grabs a screenshot. Open the incident, see what your user saw.
- AI incident explain One click sends the incident timeline to Claude. Three bullets back, plain English.
- SMS Twilio SMS for the alerts you absolutely cannot miss.
- Custom webhook templates Blade-templated webhook body. Presets for PagerDuty, Opsgenie, Mattermost.
- Embeddable widget One <script> tag drops a live status badge on your own site.
- TV mode Fullscreen status display for the team monitor. Big type, no chrome, polls every 30s.
- Multi-language Localized status pages in multiple languages. Pick your defaults, request more.
- 90/30/7-day expiry reminders Domain and SSL alerts at 90, 30, and 7 days. Configurable thresholds, no surprises.
- CSV + JSON export Pull incident and check history out as JSON or CSV any time.
- notdown CLI + manifest notdown add, cron, alert, apply. Monitoring as code in a tiny notdown.yml.
- PageSpeed + Core Web Vitals Daily PageSpeed Insights audit. Alerts when LCP, INP, or CLS slip into the red.
- Claude Code skills Drop-in skills for Claude Code, Cursor, Claude Desktop. Quickstart, triage, repo audit in one prompt.
AI to delegate everything.
Anything you'd open the dashboard for, ask your agent instead. It sees your code, your monitors, your defaults, and picks the right combination of tools to figure it out.
Works in Claude, Cursor, Codex, Copilot, pretty much any MCP-compatible client.
Free while it's new.
no card · no quotas · no quotes
Free
everything, for now$0/mo
Every monitor type, every alert channel, every MCP client.
- ✓ HTTP, cron, domain, SSL
- ✓ 60s checks
- ✓ Email, Telegram, Slack, Discord, webhooks
- ✓ Content assertions
- ✓ Maintenance windows + mute
- ✓ Status pages, subscribers, TV mode, i18n
- ✓ REST API + MCP server
- ✓ JSON + RSS feeds
Paid tier
TBD
There'll be a paid plan once it's clear what's worth charging for. Until then, everything's on the house.
- · Probably more monitors
- · Probably faster intervals
- · Probably things not yet built
FAQ
Frequently asked questions
Everything worth knowing before you set up your first monitor.
Project
Who builds Notdown?
Just me. One developer doing the design, the code, the servers, and the replies to your emails. No team, no investors, no sales department behind a curtain.
Why does Notdown exist?
I wanted simple monitoring for my own projects and kept bouncing between tools that either cost too much, split basic features across paid tiers, or were built for ops teams I don't have. So I started building the one I'd actually want to use.
What stage is the product in?
Honestly, early. Monitoring works and runs against real endpoints, but a lot of the product around it is still being built. The waitlist exists because I'd rather let people in slowly and actually help them than open the doors to something half-finished.
Where is Notdown based?
European Union. Built and operated from the EU, hosted on EU infrastructure, governed by EU law.
Roadmap
What's on the roadmap?
More alert channels, a polished status page, a REST API, an MCP server so AI agents can drive monitors, a CLI, and eventually team accounts. What I work on next is mostly decided by what waitlist users actually ask for.
Can I request a feature?
Please do. Reply to any email I send or just write me directly. I read everything, and requests from people on the waitlist or already using Notdown shape the roadmap more than anything else.
When does Notdown leave beta?
When it feels right. I'd rather take the time to get it stable than pick a date and ship something I'm not happy with. Pricing will show up here before anyone gets charged.
Audience
Who is Notdown for?
Solo developers, side projects, late-night hobby infra, and small teams that don't have anyone whose job is ops. If you just want to know when your stuff is down without learning a whole new product, that's the target.
Who is Notdown not for?
Big on-call rotations, enterprise NOCs, anyone who needs SOC 2, SAML SSO, or formal audit logs today. I'm not going to pretend to fit there — PagerDuty or BetterStack will serve you much better.
Does Notdown work for teams?
Not yet. Right now it's single-user accounts. Multi-seat teams are on the list, but they aren't built.