# On Call Handbook

> Be effective on-call — prepare, triage, respond, and hand off with clear procedures.

- Skill: `rahulrachhoya/on-call-handbook` (Agent Skill)
- Install (CLI): `npx skillmds@latest add rahulrachhoya/on-call-handbook`
- Raw SKILL.md: https://api.skillmd.com/api/skills/rahulrachhoya/on-call-handbook/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: RahulRachhoya (https://skillmd.com/u/rahulrachhoya)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/rahulrachhoya/on-call-handbook

---


# On-Call Handbook

## Before Your Shift
- Read recent incidents and postmortems
- Check the runbook is up to date for known issues
- Verify your access (production logs, dashboards, deployment tools)
- Know who to escalate to (your backup, senior engineers, manager)

## During Your Shift
- Respond to alerts within the SLO (usually 5-15 min for SEV-1)
- Acknowledge every alert — silence means no one is looking
- Triage first: who is affected? How badly? Is data at risk?
- Communicate: post in the incident channel with status updates
- Mitigate before fixing: stop the bleeding first

## Triage Flow
1. Is this real? (check monitoring, confirm with user reports)
2. What's the severity? (SEV-1/2/3/4)
3. Who needs to know? (team, manager, wider org)
4. What's the immediate fix? (rollback, feature flag, restart)
5. Document the timeline

## Handoff
- Write a clear handoff document: what's happening, what was tried, what's next
- Walk through the current state with the incoming on-call
- Don't hand off in the middle of an incident — stay until resolved or explicitly handed over
- Update the runbook with anything you learned