---

name: prep-meetings description: Prepare for customer meetings by aggregating context from multiple data sources. Invoked with /prep-meetings.

Customer Meeting Prep

Generate concise meeting prep summaries by pulling context from multiple data sources.

Invocation

• /prep-meetings - Prep for all customer meetings on today's calendar
• /prep-meetings <CustomerName> - Prep for a specific customer

Data Sources

This skill aggregates data from:

1. Vitally - Account health, notes, CSM assignment
2. Granola - Past meeting notes and transcripts
3. Slack - Recent channel activity
4. Linear - Open issues and feature requests
5. Zendesk - Support tickets
6. Grafana - Product usage metrics

Execution Flow

Step 1: Identify Customer

• From calendar: Extract domain from non-Socket attendee email
• From argument: Use provided customer name

Step 2: Match to Vitally Account

Search Vitally accounts by domain or name match.

Step 3: Fetch Data (Parallel)

Query all data sources simultaneously to minimize latency.

Step 4: Generate Summary

Output formatted prep document.

Output Format

Copy═══════════════════════════════════════════════════════════════
MEETING PREP: {Customer Name}
Meeting: {Title} @ {Time}
═══════════════════════════════════════════════════════════════

ACCOUNT OVERVIEW
├─ ARR: ${amount}
├─ Renewal: {date} ({days} days)
├─ Products: {SCA, Firewall, Extension Guard}
├─ AE: {account_executive}
└─ Health: {score}

PRODUCT USAGE (Last 30 Days)
├─ Licensed Seats: {count}
├─ Active Committers: {count} ({utilization}%)
├─ Full Scans: {count}
└─ PR Scans: {count}

ALERTS
├─ Critical: {count}
├─ High: {count}
├─ Medium: {count}
└─ Low: {count}

RECENT NOTES
• {date}: {summary}
• {date}: {summary}

SLACK HIGHLIGHTS
• {key theme}
• {key theme}

OPEN LINEAR ISSUES ({count})
• {TEAM}-{id}: {title}
• {TEAM}-{id}: {title}

SUPPORT TICKETS ({count} in 30 days)
• #{id}: {subject} ({status})

LAST MEETING
{date}: {title}
Key points: {summary}
───────────────────────────────────────────────────────────────

Implementation Notes

1. Parallel execution: Fetch from all sources simultaneously
2. Graceful degradation: Show available data even if some sources fail
3. Date formatting: Use relative dates ("3 days", "2 weeks")
4. Summarization: Keep each section to 2-3 bullet points max