Gateway Ops Observability

Observe and debug local agent gateways, daemons, health probes, logs, and usage costs.

When to Use

Use this Note when a Memoire or Studio run needs gateway logs, agent health, daemon status, usage cost, ops troubleshooting.

Workflow

1. Check daemon status, gateway health, model auth, and channel connectivity before debugging prompts.
2. Collect logs with timestamps and session ids so failures can be tied to a specific agent turn.
3. Record usage cost and quota state when provider limits or slowdowns are plausible.
4. Distinguish cached health snapshots from live probes in reports.
5. When a gateway is unhealthy, stop new tasks before restarting or changing config.

Sources