/pbx-health-check

description: Quick health check for a connected 3CX PBX
argument-hint: ""
arguments: []

PBX Health Check

A quick read-only sweep of PBX liveness: current server time, service status, and a recent-window scan of the structured event log for a failure pattern.

Prerequisites

• A 3CX PBX connected (through Conduit's vendor catalog, or directly) with

a valid, authenticated MCP session — see the `api-patterns` skill

Steps

1. **Confirm connectivity**

Get the current PBX server time. A successful response confirms the session is live and authenticated before spending calls on anything else.

2. **Check service status**

List PBX services and their state. Note any service not in a healthy running state.

3. **Scan the recent event log**

Search/list the structured PBX event log for a recent window (start with the last few hours). Look for a *pattern* of related failures rather than treating a single isolated entry as significant — PBX event logs are noisy by nature.

4. **Synthesize**

Report:

• Whether the PBX responded and server time looks correct (not wildly

skewed from expected)

• Any service not in a healthy state, named explicitly
• Any recurring event-log pattern in the recent window, with

timestamps

Do not report "PBX is healthy" as a single yes/no if any service is degraded or the event log shows a repeating pattern — state each finding explicitly so the reader can judge severity themselves.

Examples

/pbx-health-check

Error Handling

• **Server time call fails or times out:** The MCP session itself may have

expired — re-authenticate per the `api-patterns` skill before assuming the PBX itself is down.

• **Event log search returns nothing:** Confirm the queried time window is

correct before reporting "no events" — an empty result for an unexpectedly narrow or misaligned window looks identical to a genuinely quiet PBX.

Related Commands

• `/queue-status` — staffing-specific check, narrower than this general health sweep