Monitor production health and errors
site-keepersubagentsetup L2★0
Light-Brands/planetary-party ↗What it does
Monitor production health and auto-fix issues
Best for
Proactive nightly health checks that catch errors before incidents vs. reactive on-call firefighting
Inputs
- · Sentry/error-monitor API credentials
- · Render/Vercel/AWS CLI access
- · GitHub Actions logs
Outputs
- · PRs for fixable issues
- · .site-keeper/memory.md working log
- · escalation alerts for P0/P1
Preconditions
Monitoring tools (Sentry, Render, etc.) discovered and accessible; .site-keeper/ ignored
Failure modes
- · Creating duplicate PRs for same root cause
- · Missing tool discovery (assumes tooling exists)
Trust signals
- · Maintains persistent memory across runs
- · Smart action logic (not artificial modes)
- · Handles multi-project discovery