Multi-Agent Role Designer
Designs a clean division of labor for a multi-agent system, defining each sub-agent's role, boundaries, and handoff protocol so agents don't overlap or conflict.
Developers and technical founders building multi-agent systems (using tools like Claude Agent SDK, LangGraph, or CrewAI) who need a clear role architecture before writing orchestration code.
You are a multi-agent systems architect who designs clean, non-overlapping agent role structures.
Overall system goal: {{overall_system_goal}}
Workflow the agents need to cover: {{workflow_description}}
Orchestration approach (if known, e.g., a manager agent, sequential pipeline, or peer-to-peer): {{orchestration_approach}}
Do the following:
1. Propose a set of distinct sub-agents needed to cover the full workflow (aim for the minimum number that avoids overlap, typically 2-5). For each agent, define: its name/role, its single core responsibility, and what it should explicitly NOT do (to prevent scope overlap with other agents).
2. Define the handoff protocol between agents: what exact information or artifact gets passed from one agent to the next, and in what format.
3. Identify the orchestration pattern that fits best (a manager/coordinator agent directing others, a sequential pipeline, or independent agents working in parallel with a final merge step) and explain why, if not already specified.
4. Flag the single most likely coordination failure point in this design (e.g., where two agents might both think they own a task, or where a handoff could lose important context) and suggest a specific safeguard.How to use
- Describe the overall system goal and the full workflow it needs to accomplish.
- Mention your intended orchestration approach if you already have one in mind, or leave it open for a recommendation.
- Use the defined agent roles and 'explicitly NOT do' boundaries directly as scoping guidance when writing each agent's system prompt.
- Build the suggested safeguard into your orchestration logic to prevent the flagged coordination failure.
Example input
Overall goal: Automatically research, write, and fact-check blog posts for a company blog. Workflow: Topic research, draft writing, fact verification, final formatting. Orchestration approach: Not yet decided.
Example output
4 proposed agents: a Research Agent (gathers sources, does not write content), a Writer Agent (drafts the post from research, does not verify facts itself), a Fact-Checker Agent (verifies claims against sources, does not rewrite prose), and a Formatter Agent (applies final formatting, does not alter factual content), a defined handoff protocol specifying research gets passed as a structured source list, a recommendation for a sequential pipeline given the linear dependency between steps, and a flagged risk that the Fact-Checker might lose context on which claims came from which source, with a safeguard of requiring the Writer Agent to inline-tag claims with source references.
Customization tips
- — If you already have some agents built, describe them so the design only proposes new agents needed to fill genuine gaps.
- — For cost-sensitive systems, ask for a version that minimizes the number of agents and LLM calls even if it means slightly broader individual roles.
- — Request a visual description of the data flow between agents if you need to communicate the architecture to a team.
Tags
FAQ
What is this prompt for?
How should I customize it?
Are there any limitations?
How is it different from a basic prompt?
Related Prompts
All AI Agents prompts →Agent Task Decomposition Planner
Breaks a broad goal into a structured set of sub-tasks an AI agent can execute step by step, with clear success criteria and handoff points for human review.
Agent Handoff Contract Designer
Define reliable handoffs between agents or from agents to people, with clear payloads, acceptance rules, ownership, and recovery paths.
Agent System Prompt Stress Tester
Pressure-tests an AI agent's system prompt by generating edge-case user inputs designed to expose ambiguity, scope creep, or unsafe behavior before deployment.
Related Skills
AI Agent Prompt & Tool Spec Builder
Designs a complete AI agent specification — system prompt, tool definitions, and decision boundaries — from a description of the task you want the agent to handle.
Multi Agent Collaboration Orchestrator
Design multi-agent workflows with roles, task decomposition, shared state, handoffs, verification, budgets, conflict handling, and human control.
Agent Prompt & Tool Spec Designer
Designs a complete system prompt and tool specification for an LLM agent from a description of what the agent should do.
Related Articles
How do AI agents use tools, memory, and APIs?
Agents use tools and APIs to retrieve data or act in external systems, while memory helps preserve selected context. Both need permission limits and review design.
How is an AI agent different from a chatbot?
A chatbot usually answers messages. An AI agent can plan, use tools, follow a workflow, make decisions within boundaries, and sometimes take actions after approval.
What is an AI agent in simple words?
An AI agent is a system that uses a model to reason through a task and may use tools, memory, or workflows to complete steps beyond a single chat response.