canva-incident-runbook

name	canva-incident-runbook
description	Execute Canva Connect API incident response with triage, mitigation, and postmortem. Use when responding to Canva-related outages, investigating API errors, or running post-incident reviews for Canva integration failures. Trigger with phrases like "canva incident", "canva outage", "canva down", "canva on-call", "canva emergency", "canva broken".
allowed-tools	Read, Grep, Bash(kubectl:*), Bash(curl:*)
version	1.0.0
license	MIT
author	Jeremy Longshore <[e-mail]>
tags	["saas","design","canva"]
compatible-with	claude-code

Canva Incident Runbook

Overview

Rapid incident response for Canva Connect API integration failures. Covers triage, mitigation, escalation, and postmortem.

Quick Triage (First 5 Minutes)

#!/bin/bash
# canva-triage.sh — Run immediately when incident detected

echo "=== Canva Triage ==="

# 1. Is it Canva or us?
echo -n "Canva API: "
curl -s -o /dev/null -w "HTTP %{http_code} (%{time_total}s)\n" \
  -H "Authorization: Bearer $CANVA_ACCESS_TOKEN" \
  "https://api.canva.com/rest/v1/users/me"

# 2. Check our health endpoint
echo -n "Our health: "
curl -s -o /dev/null -w "HTTP %{http_code}\n" \
  "https://api.ourapp.com/health"

# 3. Error rate (if Prometheus available)
echo "Error rate (5min):"
curl -s "localhost:9090/api/v1/query?query=rate(canva_api_errors_total[5m])" \
  | python3 -c "import sys,json; d=json.load(sys.stdin); print(d['data']['result'])" 2>/dev/null \
  || echo "Prometheus not available"

# 4. Rate limit status
echo -n "Rate limit remaining: "
curl -sD - -o /dev/null -H "Authorization: Bearer $CANVA_ACCESS_TOKEN" \
  "https://api.canva.com/rest/v1/designs?limit=1" 2>&1 \
  | grep -i "x-ratelimit-remaining" || echo "unknown"

Decision Tree

API returning errors?
├── YES → What HTTP status?
│   ├── 401 → Token expired → Refresh token, check rotation
│   ├── 403 → Scope issue → Verify integration permissions
│   ├── 429 → Rate limited → Enable backoff, check Retry-After
│   ├── 5xx → Canva outage → Enable fallback, monitor status page
│   └── Other → Check request format against API docs
└── NO → Is our integration healthy?
    ├── YES → Likely resolved or intermittent → Monitor
    └── NO → Check our infra (pods, memory, DNS, TLS)