Step 7 Monitor Setup

Step 7 · Monitor Setup

Command: /monitor-setup [project] · Habit: H7 Sharpen the Saw · Previous: Step 6 · Deploy Guide

Purpose

Set up error tracking, alerting, and health checks so you learn from production. The final step closes the loop between deploy and the next planning cycle.

When to use

• After a new service is deployed
• When adding a new critical feature to an existing service
• When an incident reveals a monitoring gap

When to skip

• Internal tools with zero users (use judgment — often you still want basic health)
• Ephemeral preview deploys

Process

1. Read existing monitoring — don't duplicate what's already wired up
2. Define the 4 golden signals: latency, traffic, errors, saturation
3. Add or verify:
   • /health endpoint (liveness + readiness)
   • Structured logging with correlation IDs
   • Error tracking (Sentry, Rollbar, etc.)
   • Metrics export (Prometheus, OpenTelemetry)
   • Alerts on SLO breach
4. Document the runbook: "If X alerts fire, do Y"

Output

• Monitoring config (code + infra)
• Alert rules
• Runbook entry for this feature
• Dashboard link

Handoff

• Expects: A deployed service from /deploy-guide
• Produces for the next cycle: Production feedback that informs the next /research or /requirements

H7 Checkpoint

"Am I investing in future capability, or just grinding out output?"

See also

• Source: skills/monitor-setup/SKILL.md
• Habits Reference → H7
• Next cycle: Step 0 · Research