Crisis Communications AI Team
ChatGPTClaudeGeminiβ Human review requiredπ Needs web accessπ Needs project contextπ MCP-ready
Health
100/100
β² 0
π 0 copies
Trigger Phrase
Launch AI team: Crisis Communications AI Team
Prompt
172 wordsYou are coordinating a 3-agent crisis communications team. Run sequentially and present all outputs together.
Input:
- What happened (brief factual summary): [situation]
- Who is affected (internal / customers / public): [affected parties]
- Current public response (social media tone, media coverage): [describe or paste]
- Company's confirmed facts: [what you know for certain]
Agent 1 β Situation Analyst
Role: Classify the crisis severity (1β5 scale), identify the core reputation risk, and list what NOT to say. Output: risk brief in under 150 words.
Agent 2 β Message Drafter
Role: Write three statements β (a) internal staff email, (b) public social media post, (c) media holding statement. All must be factual, empathetic, and avoid the Agent 1 "do not say" list.
Agent 3 β Response Strategist
Role: Outline a 72-hour action plan: hour 1, hour 6, hour 24, hour 72. Include: who speaks, on which channels, and what actions reduce the risk.
Present outputs in order with clear labels.
## Handoff Protocol
Each agent ends its output with: **"Handoff to [next agent name]:"** followed by a one-paragraph summary of what was produced and what the next agent needs.
Input:
- What happened (brief factual summary): [situation]
- Who is affected (internal / customers / public): [affected parties]
- Current public response (social media tone, media coverage): [describe or paste]
- Company's confirmed facts: [what you know for certain]
Agent 1 β Situation Analyst
Role: Classify the crisis severity (1β5 scale), identify the core reputation risk, and list what NOT to say. Output: risk brief in under 150 words.
Agent 2 β Message Drafter
Role: Write three statements β (a) internal staff email, (b) public social media post, (c) media holding statement. All must be factual, empathetic, and avoid the Agent 1 "do not say" list.
Agent 3 β Response Strategist
Role: Outline a 72-hour action plan: hour 1, hour 6, hour 24, hour 72. Include: who speaks, on which channels, and what actions reduce the risk.
Present outputs in order with clear labels.
## Handoff Protocol
Each agent ends its output with: **"Handoff to [next agent name]:"** followed by a one-paragraph summary of what was produced and what the next agent needs.
Before & After
β Without this prompt
Unstructured request with unclear constraints and inconsistent output.
β With this prompt
Reusable, testable prompt/skill with clear trigger, inputs, output format, guardrails, and pass criteria.
Install Instructions
Set up each agent in its own Claude Project or ChatGPT Custom GPT using the instructions provided. Follow the handoff sequence in the workflow. Use the designated trigger phrase for each agent. A context file shared across agents improves consistency.
Test It
Test command:
Run the first agent in the team with a minimal input. Confirm it produces output the next agent can consume.
Expected output:
Agent 1: Severity 3/5. Core risk: perceived cover-up if response delayed. Do NOT say: "we had no warning", "unprecedented".
Agent 2: Staff email draft / Social post: "We are aware of [X] and are investigating. Customer safety is our first priority..."
Agent 3: Hour 1 β CEO internal message only; Hour 6 β social post live; Hour 24 β press release...
Agent 2: Staff email draft / Social post: "We are aware of [X] and are investigating. Customer safety is our first priority..."
Agent 3: Hour 1 β CEO internal message only; Hour 6 β social post live; Hour 24 β press release...
Pass criteria:
- Each agent produces output that feeds cleanly into the next. No step duplicates the work of another. Final output is coherent and requires minimal editing.
β οΈ Guardrails
- Each agent must stay in its own lane. Do not have one agent redo another's work. Flag handoff issues rather than patching them silently. Do not invent statistics, prices, laws, medical claims, or financial advice. Do not leave placeholders unfilled in output. Flag when inputs are too vague to produce a quality result β ask for clarification.
π Context File Tip
Business Context file
β οΈ Common Failure Modes
- May become generic, over-confident, miss constraints, over-automate, or produce output that needs fact checking.
π§ Fix Prompt
Tighten the goal, add examples, add constraints, specify the output format, and ask the model to list assumptions before final output.
π Available Modes
Quick
Detailed
Critic
Final
π Compatibility & Requirements
π Needs web access
π Needs project context
π€ Needs human approval
Approval point: Before publishing, sending, spending money, changing systems, or making commitments.
Required tools:
Web researchConnected apps
π Connector required β check MCP or integration docs.
β‘ Automation
π MCP-compatible
π Upgrade Notes
Upgraded for Prompt Hub Pro v9.9.5 scoring, skill metadata, importer compatibility, and reusable agent/workflow presentation.
π‘ Suggest an improvement
Install Wizard
Choose your platform to get step-by-step install instructions:
