IR-003 critical on-call-escalation
Contact list for emergencies
During an incident, you shouldn't be hunting for phone numbers. A readily accessible contact list with multiple reach methods saves critical minutes.
Question to ask
"If Slack is down, how does your team reach each other?"
Pass criteria
- ✓ Contact list exists with key people
- ✓ Multiple contact methods per person (phone + Slack/email)
- ✓ Accessible during outages (not solely in a system that could be down)
Fail criteria
- ✗ No contact list
- ✗ List exists but only has email
- ✗ List is only in production systems
Related items
Verification guide
Severity: Critical
During an incident, you shouldn't be hunting for phone numbers. A readily accessible contact list with multiple reach methods saves critical minutes.
Check automatically:
- Look for contact documentation:
# Search for contact lists
grep -riE "contact|emergency|phone|mobile|\+1|\+44" docs/ runbooks/ README.md CLAUDE.md CONTACTS.md ONCALL.md --include="*.md" 2>/dev/null
# Look for dedicated contact files
find . -maxdepth 3 -name "*contact*" -o -name "*emergency*" -o -name "*oncall*" 2>/dev/null | grep -v node_modules
Ask user:
- "Where's the emergency contact list stored?"
- "Does it include multiple contact methods (phone, Slack, email)?"
- "Is it accessible when your primary systems are down?" (not just in a wiki that might be offline)
Cross-reference with:
- IR-001/IR-002 (on-call and escalation) - contact list supports both
- IR-004 (incident management tool) - tool often stores contacts
- Section 12 (monitoring) - alerting tools need contact info configured
Pass criteria:
- Contact list exists with key people (on-call, escalation contacts, vendors)
- Multiple contact methods per person (phone + Slack/email)
- Accessible during outages (not solely in a system that could be down)
Fail criteria:
- No contact list
- List exists but only has email (useless at 3am)
- List is only in production systems (inaccessible during outage)
Evidence to capture:
- Location of contact list
- Contact methods included (phone, Slack, email, etc.)
- Whether it's accessible offline/during outages