Skip to main content

Incident Management Procedure (A.5.24)

Document ID: incident-management-procedure
Version: 1.0
Effective from: 2025-01-01
Owner: Founder / CEO
Review cycle: Annual (or after any Sev1 incident)


Incident classification

SeverityDefinitionResponse SLA
Sev1Active breach, data exfiltration, key compromise, production outage affecting customersImmediate — respond within 15 min
Sev2Suspected breach, significant degradation, external report of vulnerabilityRespond within 2 hours
Sev3Minor degradation, isolated control failure, stale evidence gapRespond within 24 hours
Sev4Near-miss, policy deviation with no immediate riskRespond within 72 hours

Detection

Incidents may be detected via:

  • Automated gap alerts from collectors (Discord webhook)
  • cron-health flagging stale evidence
  • External vulnerability disclosure
  • Customer report via support portal
  • Cloudflare security alerts
  • GitHub Advanced Security alerts

Response procedure

Step 1 — Detect and log

  1. Record the incident in the compliance portal (/compliance/risks → create incident)
  2. Classify severity (Sev1–4)
  3. Set status to open

Step 2 — Contain

For Sev1/2:

  • Rotate any suspect credentials immediately (rotate via Wrangler Secrets Store)
  • Revoke affected access (GitHub org, Cloudflare account)
  • If signing key compromised: set state to revoked in signing_keys, generate new key
  • Enable IP allowlist break-glass if needed

Step 3 — Investigate

  • Correlate with evidence_runs and audit_events
  • Pull relevant evidence from R2 for the affected time window
  • Determine root cause

Step 4 — Eradicate and recover

  • Deploy fix
  • Verify evidence collection resumes normally
  • Run full cron trigger to regenerate any missing evidence

Step 5 — Post-incident review

  • For Sev1/2: write postmortem within 5 business days
  • Upload postmortem to R2, link in incident record
  • Review whether controls need strengthening

Breach notification

If a personal data breach affects EU residents (GDPR Art. 33): notify the relevant supervisory authority within 72 hours. If the breach is likely to result in high risk to individuals (Art. 34): notify affected data subjects without undue delay.

Evidence

CollectorCadenceControl
incident-logMonthlyA.5.24–A.5.28
cron-healthDailyA.5.36, A.8.16