Building a central guardrail dashboard for AI agents and cron jobs, would love feedback

Hey PH,

I’m building a tool that centralizes retry limits, spend caps, and alerts across every cron job or agent you run, instead of wiring guardrails into each project separately.

The idea came from noticing how often solo devs and small teams only find out an agent looped or retried for hours after the bill shows up. Most people patch this per project (a retry cap here, a budget alert there), but there’s nowhere to see it all in one place or stop it before it gets expensive.

Still early and heads down on the build. If you’re running AI agents or cron jobs in production, especially unattended ones, I’d love to hear how you’re handling this today, and whether you’d be up for trying an early version when it’s ready.

Happy to answer questions below.