I built an uptime monitor for web agencies – here's what I learned
I run a 3-client web dev agency on the side of my day job. One of my clients' sites went down and I didn't know about it until they called me. Embarrassing.
So I built Client Uptime (https://clientuptime.com) to solve my own problem: dead-simple uptime monitoring designed for agencies managing multiple client sites.
What it does:
- Monitor HTTP/HTTPS endpoints (1-min checks)
- Email + Slack alerts on downtime
- Branded public status pages (share with clients to look professional)
- Client management (group monitors by client)
- Incident tracking
- Webhooks + API keys
- MCP server for AI agent integration
Pricing: Free tier with 50 monitors (same as Uptime Robot). $5/mo for 500 monitors (cheaper than Uptime Robot).
The interesting part: I added an MCP server so AI agents can control the whole thing. My AI assistant can now create monitors when I onboard a new client, check on outages, and pull uptime reports. It's like having a junior DevOps person who never sleeps.
Stack: Node.js, SQLite, runs on a single VPS. Total hosting cost: about $6/mo. 50 free monitors matches Uptime Robot's free tier, and the paid plan at $5/mo gives you 10x that.
Revenue so far: $0. Just launched. Looking for feedback and early users.
I'm curious what other indie hackers think about the MCP/AI-agent angle. Is that something you'd actually use, or is it a gimmick?
