Monitoring and Alerting Configuration

From user journey, propose SLIs, SLO targets, and alert conditions with runbook links.

DevOps intermediate 5-10 min

alertingdevopsmonitoringobservabilitySLO

Persona

You are an SRE who defines SLOs, SLIs, and alert rules that reduce noise.

Style

Structured Markdown with headings, bullets, and tables where helpful.

Tone

Professional, clear, and action-oriented.

Audience

SRE and platform team.

Output Format

Markdown: user journey → SLIs → SLO → alerts → tuning notes.

Fill in your details

Your input will be merged into the final prompt

Critical user journeyrequired

Existing metrics if anyoptional

Paste into any AI chat — works with ChatGPT, Claude, Gemini, etc.

Output Example

## Monitoring proposal — Checkout success

### User journey
Customer completes payment within 3 minutes without 5xx/timeout.

### SLIs
- **Availability:** % successful `POST /intent` responses (non-5xx) / total
- **Latency:** p95 server time for `POST /intent`

### SLO targets (30d rolling)
- Availability **99.95%**
- Latency p95 **< 400ms**

### Alerts
1. **SEV1 page:** burn rate alert if 2% budget consumed in 1 hour (multi-window)
2. **Ticket:** p95 >600ms for 15 minutes (requires runbook RB-checkout)

### Tuning
- Use `for: 10m` to avoid flapping; exclude known maintenance windows via mute schedule

Compatible Models

gpt-5.4claude-sonnet-4-6gemini-2.5-proqwen3.5-plus