Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →**Feature**: 042-twilio-voice-mcp
.claude/skills/automateyournetwork-twilio-outbound-call-skill/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 125% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 282% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 27% | 0% |
Feature: 042-twilio-voice-mcp Priority: P2 User Stories: US2 (On-Demand), US3 (Situation Updates)
Enable John to request on-demand phone calls for network status updates and schedule periodic update calls during ongoing incidents.
This skill activates when:
User Request → "Call me with network status"
│
├─► Gather Network Status
│ └─► Query relevant MCP servers
│
├─► Generate Voice Summary
│ └─► Concise status overview
│
├─► Pre-call Validation
│ ├─► Rate limit check
│ ├─► Quiet hours check
│ └─► Whitelist validation
│
├─► Sanitize Content
│ └─► Remove IPs, credentials
│
└─► Initiate Call via TwilioSchedule Request → "Call me every 30 minutes"
│
├─► Register Schedule in Memory MCP
│ └─► Store interval, incident_id
│
├─► [Loop] On Schedule Trigger
│ │
│ ├─► Check Incident Status
│ │ └─► Resolved? → Cancel schedule
│ │
│ ├─► Check for Changes
│ │ └─► No updates? → Skip call
│ │
│ └─► Initiate Status Call
│
└─► Critical Change Detection
└─► Immediate call on: resolved, escalatedjson{ "tool": "twilio_voice_call", "arguments": { "to": "[WHITELIST_NUMBER]", "message": "Network status update. All 47 devices are operational. No active incidents. BGP sessions healthy.", "priority": "normal" } }
json{ "tool": "twilio_voice_check_rate_limit", "arguments": {} }
json{ "tool": "twilio_voice_check_quiet_hours", "arguments": { "priority": "normal" } }
Network status update.
{device_count} devices monitored.
{healthy_count} healthy, {warning_count} warnings, {critical_count} critical.
{active_incidents} active incidents.
Most recent: {recent_event}.Status for {device_name}.
Device type: {device_type}.
State: {operational_state}.
CPU: {cpu_percent}%, Memory: {mem_percent}%.
Uptime: {uptime}.Update on incident {incident_id}.
Status: {status}.
Duration: {duration}.
Affected: {affected_services}.
Latest update: {update_summary}.json{ "scheduled_calls": { "max_interval_minutes": 60, "min_interval_minutes": 15, "auto_cancel_on_resolve": true, "skip_if_no_changes": true } }
User: "Call me with the network status"
Action:
1. Check rate limits (OK)
2. Check quiet hours (OK)
3. Gather status from MCP servers
4. Generate summary: "47 devices operational, no incidents"
5. Sanitize content
6. Initiate call to [WHITELIST_NUMBER]User: "Call me every 30 minutes about incident INC12345"
Action:
1. Register schedule in Memory MCP
2. Every 30 minutes:
a. Check if incident resolved → Cancel if yes
b. Check for status changes → Skip if none
c. Generate update message
d. Initiate call
3. On incident resolve → Final call + cancel scheduleOther measured skills in the registry, with their headline benchmark lift.