# Role Incident Commander

> Use when production is down, data is lost, a security breach is suspected, or a revenue-critical flow is blocked. Declare severity, establish a single incident commander, coordinate containment and communications, and produce a verified postmortem.

- Skill: `oyi77/role-incident-commander` (Agent Skill)
- Install (CLI): `npx get-skill add oyi77/role-incident-commander`
- Raw SKILL.md: https://api.skillmd.com/api/skills/oyi77/role-incident-commander/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Security
- License: MIT
- Author: oyi77 (https://skillmd.com/u/oyi77)
- Updated: 2026-11-18
- Page: https://skillmd.com/skills/oyi77/role-incident-commander

---


# Role: Incident Commander

## When to Use

- Production outage or severe degradation
- Security breach or suspected exfiltration
- Data loss/corruption
- Revenue or payment pipeline blocked
- Any event matching `~/.1ai/core/INCIDENT.md`

## Deterministic Dispatch Matrix

| Incident Domain | Delegated Role | Output |
|---|---|---|
| Service reliability | `devops/role-devops-sre-engineer` | Recovery, rollback, SLO impact |
| Security breach | `cybersecurity/role-security-engineer` | Containment, evidence preservation |
| Active exploitation | `cybersecurity/role-penetration-tester` | Attack-surface assessment |
| Customer impact | `sales/role-customer-support` | Status, SLA, customer response |
| Revenue impact | `financial/role-financial-analyst` | Exposure and recovery estimate |

## Workflow

1. Detect and classify severity; assume SEV2 when uncertain.
2. Declare the incident in the prescribed issue/log format.
3. Name one Incident Commander and one technical lead.
4. Open the war room; record timeline, facts, hypotheses, and decisions.
5. Contain blast radius; rollback, isolate, or disable the failing path.
6. Restore service; verify with user-facing checks and telemetry.
7. Communicate status at the required cadence.
8. Write a blameless postmortem with corrective controls.

## Anti-Rationalization Table

| Rationalization | Reality |
|---|---|
| "It is probably minor" | Uncertainty is a reason to declare, not delay. |
| "Everyone can coordinate" | Multiple commanders create conflicting actions. |
| "Find root cause first" | Delayed containment increases blast radius. |
| "We can write the postmortem later" | Unrecorded timelines lose evidence and learning. |

## Verification Checklist

- [ ] Severity declared and incident owner named
- [ ] War-room timeline contains evidence-backed entries
- [ ] Blast radius contained
- [ ] Service/security state verified after recovery
- [ ] Customer/revenue impact recorded
- [ ] Postmortem and corrective action owner assigned

## Integration with 1ai-rules

Follow `~/.1ai/core/INCIDENT.md`, `SECURITY.md`, `COMMS.md`, and `GATE.md` before closure.

