Back to Cookbook

Shift Kickoff

Start your on-call shift knowing what's on fire, not hunting for context

Generates a shift briefing at the start of your on-call rotation: recent deployments, open incidents, current alert state, and anything the previous shift flagged. Also builds instant context when you get paged for an unfamiliar service.

House RecipeWork5 min

INGREDIENTS

💬Slack✈️Telegram🐙GitHub

PROMPT

Create a skill called "Shift Kickoff". Generate on-call context: Mode 1 — Shift Briefing: At the start of my on-call rotation, gather: 1. Recent deployments from the last 24-48h (git log, CI/CD records) 2. Currently firing alerts and their duration 3. Open incidents and their status 4. Any handoff notes from the previous shift 5. Upcoming scheduled maintenance or changes Present as a concise briefing I can read in 2 minutes. Mode 2 — Incident Context: When I get paged for a specific service: 1. What the service does and who owns it 2. Service dependencies (upstream and downstream) 3. Recent changes to the service or its dependencies 4. Known failure modes and relevant runbooks 5. Current metrics: is the service healthy, degraded, or down? Present as rapid-fire bullet points for quick consumption during an incident. Mode 3 — Handoff: At the end of my shift, generate handoff notes covering ongoing issues, actions taken, and anything the next person should know.

How It Works

Starting an on-call shift or getting paged for a service you don't know

means spending 15-45 minutes gathering context before you can even start

debugging. This skill does that in seconds.

What You Get

  • Shift briefing: recent deploys, open incidents, active alerts, flagged issues
  • Service context: what the service does, its dependencies, common failure modes, and runbooks
  • Recent change log: git commits, CI/CD deployments, infrastructure changes in the last 24-48 hours
  • Alert summary: what's firing, how long, and whether it's been acknowledged
  • Handoff notes from previous shift (if available)

Setup Steps

  1. Configure your Claw with access to your deployment tools, monitoring, and git repos
  2. Run at the start of each on-call shift for a briefing
  3. When paged, give your Claw the service name for instant context
  4. At shift end, ask it to generate handoff notes for the next rotation

Tips

  • Most useful when it has access to both git history and monitoring APIs
  • The "instant context" mode is a lifesaver for pages about unfamiliar services
  • Include a list of services and their owners in your Claw's knowledge for faster lookups
  • Generate handoff notes at the end of every shift to build institutional memory
Tags:#on-call#incident-response#devops#automation