monitoring

name	monitoring
description	Monitors key metrics and sends alerts for revenue, usage, enterprise accounts, backend costs, and API runtime. Routes requests to specialized monitoring sub-agents.
tags	["monitoring","alerts","metrics"]

Monitoring Agent

When to use

• "Monitor revenue metrics"
• "Set up usage alerts"
• "Track enterprise account health"
• "Monitor backend costs"
• "Alert on API runtime issues"
• "Create monitoring dashboard"
• "Set up anomaly detection"

What it does

Routes monitoring requests to specialized sub-agents based on the monitoring type:

Monitoring Type	Sub-Agent	File
Revenue tracking	Revenue Monitor	agents/monitoring/revenue/SKILL.md
Feature/user usage	Usage Monitor	agents/monitoring/usage/SKILL.md
Enterprise accounts	Enterprise Monitor	agents/monitoring/enterprise/SKILL.md
Backend/GPU costs	BE Cost Monitor	agents/monitoring/be-cost/SKILL.md
API runtime & performance	API Runtime Monitor	agents/monitoring/api-runtime/SKILL.md

How to route

1. Identify monitoring type from user request
2. Read the appropriate sub-agent SKILL.md
3. Follow that sub-agent's workflow

Examples

User says...	Route to...
"Alert me when revenue drops"	Revenue Monitor
"Track feature adoption rates"	Usage Monitor
"Monitor McCann account usage"	Enterprise Monitor
"Alert on GPU cost spikes"	BE Cost Monitor
"Track API latency issues"	API Runtime Monitor

Reference files

All monitoring agents use:

File	Read when
shared/bq-schema.md	Before writing SQL
shared/metric-standards.md	Before defining metrics
shared/event-registry.yaml	Before referencing events

Rules

• DO identify the monitoring type before routing
• DO read the sub-agent's SKILL.md completely before executing
• DO NOT mix multiple monitoring types in one workflow — route separately
• DO validate threshold values with the user before setting alerts