Skip to main content

Restarting a Service (Admins)

Sometimes a service (like the Brain Memory) needs a quick restart — for example after an update. This is safe, takes seconds, and does not lose data.

The one safe way​

Open PowerShell and run:

pwsh -NoProfile -File B:\GRIFF_WARM\AI_AGENT_PLATFORM\00_MISSION_CONTROL\service-ops\Restart-GriffService.ps1 -Action restart -Service all

Use Brain-MCP or Brain-HTTP only when you intentionally want one service. all is the normal Brain restart.

What "good" looks like​

You should see a small table ending with:

  • taskState: Running
  • healthy: True
  • taskActionOk: True
  • manager: schtask

That means it worked. A receipt file is saved automatically for the record.

Check status without restarting​

pwsh -NoProfile -File B:\GRIFF_WARM\AI_AGENT_PLATFORM\00_MISSION_CONTROL\service-ops\Restart-GriffService.ps1 -Action status -Service all

Important​

  • Do not restart Brain outside the service-ops command above; it handles task state, port cleanup, health checks, and receipts together.
  • You do not need to restart the Cloudflare tunnel — it reconnects on its own.
  • If it doesn't come back healthy, see Troubleshooting or Support.

(Technical detail for operators: Developer → Restart runbook.)