Ops center for all my projects. What should it watch next: session replays, uptime, or security?

by•

Over the last year I went from one side project to 7+. Shipping got cheap; keeping an eye on everything didn't. So I built the screen I wanted: every project in one place, sorted by what needs me first.

Today it covers, for every project from one script tag:

  • Traffic: live visitors, sessions, sources and funnels, cookieless by default

  • Errors: browser errors, plus server errors from Node, Bun or Deno with a small package

  • Feedback: a widget where users can leave a note with a screenshot

  • Consent: a cookie banner and warnings when a site sets cookies it shouldn't

  • One "Needs attention" inbox across the whole portfolio

  • Ask it from your AI chat over MCP: "which project needs me today, and why?"

  • Scheduled status every morning over MCP: "Claude directly generates PRs with fixes if there was an error"

Now I'm stuck on what to build next. Three candidates:

1. Session replays. Watch what a user actually did before the error or the angry feedback. Most useful for "it's broken" with no details. The catch: it's heavy and it's the most privacy-sensitive data you can collect, and right now the script is small and cookieless. I'd only do it with masking on by default.

2. Health monitor. Pinging your sites and endpoints from outside and telling you when one is down or slow. Today, no events from a project only means no events. It doesn't prove the site is up. This would close that gap.

3. Security scanner. A regular check of each project for the things that bite fast-shipped apps: API keys exposed in the frontend bundle, missing security headers, open database tables, expired certificates.

If you run more than one project:

  1. Which of these would you actually use every week?

  2. Which one has burned you before? A site down for days, a leaked key, a bug report you couldn't reproduce?

  3. Is there something missing that I haven't listed?

What slows you down most today when launching a new project? Looking forward to any ideas and feedback.

4 views

Add a comment

Replies

Best

I would pick the health monitor and let each project define what a healthy response contains

A homepage can return 200 while the API or database connection is broken. A small read-only check of the main dependency would make the alert much more useful