Speak Incident Runbook

Overview

Rapid incident response procedures for Speak language learning-related outages.

Prerequisites

• Access to Speak dashboard and status page
• kubectl access to production cluster
• Prometheus/Grafana access
• Communication channels (Slack, PagerDuty)

Instructions

1. Severity Levels
2. Quick Triage
3. Decision Tree
4. Immediate Actions by Error Type
5. Communication Templates
6. Fallback Modes
7. Post-Incident

For full implementation details, load: Read(${CLAUDE_SKILL_DIR}/references/implementation-guide.md)

Output

• Issue identified and categorized
• Mitigation applied
• Stakeholders notified
• Evidence collected for postmortem
• Fallback modes enabled if needed

Error Handling

Examples

One-Line Health Check

set -euo pipefail
curl -sf https://api.yourapp.com/health | jq '.services.speak.status' || echo "UNHEALTHY"

Quick Fallback Toggle

set -euo pipefail
# Enable fallback
kubectl set env deployment/speak-integration SPEAK_FALLBACK_MODE=true

# Disable fallback (restore normal)
kubectl set env deployment/speak-integration SPEAK_FALLBACK_MODE-

Resources

• Speak Status Page
• Speak Support
• Internal Runbook Wiki

Next Steps

For data handling, see speak-data-handling.