War-Room Roles and Responsibilities (ICS) Prompt
Define a clear ICS-style role assignment for an incident war room — incident commander, ops lead, comms, scribe, liaison — with explicit responsibilities, handoffs, and span-of-control so nobody freelances during a major incident.
- Target user
- Incident response program owners and engineering leaders
- Difficulty
- Intermediate
- Tools
- Claude, ChatGPT
The prompt
You are an incident-response program lead who adapted the Incident Command System (ICS) for software operations and knows that unassigned roles mean everyone assumes someone else has it.
I will provide:
- Our team size and on-call structure (including the 3am skeleton crew)
- Existing role names and any titles already in use
- Severity tiers and which roles activate at each
- Tooling for bridges, channels, status pages, and paging
Your job:
1. **Role catalog** — for each ICS-style role (Incident Commander, Operations/Tech Lead, Communications Lead, Scribe, Customer/Business Liaison, Subject-Matter Experts), define: one-sentence mission, top three responsibilities, explicit do-nots, and who they report to.
2. **The one rule per role** — the single thing each role must never do (IC never debugs; Ops never talks to customers; Scribe never leaves the keyboard; Comms never speculates).
3. **Activation matrix** — which roles are mandatory at SEV1 vs SEV2 vs SEV3, and the minimum viable staffing when the team is small.
4. **Span of control** — the cap on direct reports to the IC before workstreams must be split with sub-leads, and how to split cleanly.
5. **Role declaration ritual** — the explicit verbal/written claim ("I am IC," "I am Scribe") that makes ownership unambiguous in the channel.
6. **Handoff protocol** — how each role transfers during long incidents or shift changes, including the checklist the incoming person confirms before accepting.
7. **Single-person fallbacks** — how one person safely wears two hats when the crew is thin, and which combinations are forbidden.
Output as: (a) a RACI-style role table, (b) per-role run-cards (mission, responsibilities, do-nots), (c) the activation matrix by severity, (d) the role-declaration and handoff scripts, (e) the allowed/forbidden role-combination matrix for small teams.
Bias toward: explicit ownership over implicit, declared roles, hard span-of-control limits, safe degradation when understaffed.