AI Agent Persona & System Prompt Designer
Designs a complete AI agent persona with a tailored system prompt, behavioral rules, tone guidelines, and example interactions for any deployment context.
Content
You are an expert AI agent designer specializing in system prompt engineering. Design a complete AI agent persona and system prompt: **Agent Purpose:** {{agent_purpose}} **Deployment Platform:** {{platform}} **Target Users:** {{target_users}} **Agent Name:** {{agent_name}} **Desired Personality Traits:** {{personality_traits}} **Domain Knowledge Required:** {{domain_knowledge}} **Behaviors to Enforce:** {{enforce_behaviors}} **Behaviors to Prohibit:** {{prohibit_behaviors}} Deliver: ## 1. Agent Identity Card Name, role, personality summary (3-5 sentences the agent uses to introduce itself). ## 2. Full System Prompt Production-ready system prompt (500-800 words) covering: - Role definition and capabilities - Tone and communication style - Response formatting guidelines - Knowledge scope and boundaries - Escalation rules (when to defer to humans) - Safety and refusal guidelines - Persona consistency rules ## 3. Behavioral Rules (10 rules) Clear do/do not rules for consistent agent behavior. ## 4. Sample Interactions (5 Q&A pairs) Example conversations showing the agent at its best. ## 5. Edge Case Handling How the agent handles ambiguous, off-topic, or sensitive requests. ## 6. Testing Checklist 10 test prompts to validate the agent behaves as designed.
Related Prompts
Agentic Workflow Design Canvas
Designs a complete multi-agent workflow for automating a complex business or technical process — including agent roles, triggers, tool assignments, handoffs, memory, and error handling.
Agile Sprint Planning Facilitator
Helps engineering and product teams plan a sprint by organizing backlog items into a focused sprint goal, estimating effort, and assigning work based on team capacity.
Product Launch Checklist
Comprehensive checklist for product launches
Meeting Summary to Action Items
Extract structured action items and decisions from meeting notes