返回提示词库
文本 · 通用大模型多Agent协作方案自然语言需求转架构设计PW
创作者Prompt2 编辑部PromptWhisper 收录
文本通用大模型开发与工程

多Agent协作方案自然语言需求转架构设计

用一句话描述你的多Agent系统需求,自动生成完整的架构设计方案,包含Agent角色定义、通信协议、任务编排和错误处理策略

11浏览
完整提示词可替换花括号中的变量后直接使用

You are a senior AI systems architect specializing in multi-agent systems. I will describe my use case in one sentence. Please design a complete multi-agent architecture: ## Multi-Agent Architecture Design ### 1. Agent Roster For each agent, specify: - **Name & Role**: What it does - **Model recommendation**: Which LLM fits best (cost vs capability) - **Tools/APIs**: What external tools it needs - **Memory**: What context it must retain ### 2. Communication Protocol - Message format (structured JSON schema) - Routing rules: who talks to whom, when - Handoff conditions and escalation paths ### 3. Orchestration Pattern - Choose: Sequential / Parallel / Hierarchical / Swarm - Justify the choice for this use case - Draw the flow as a Mermaid diagram ### 4. Error Handling & Recovery - What happens when an agent fails? - Retry strategy with backoff - Fallback agents or degraded-mode behavior ### 5. Observability - Key metrics to track - Logging strategy for debugging - Cost monitoring approach ### 6. Implementation Skeleton Provide starter code (Python) using a popular framework (OpenAI Agents SDK, LangGraph, or CrewAI). My use case: [DESCRIBE IN ONE SENTENCE]

2026/4/24

如何使用这条提示词

  1. 1复制上方完整提示词。
  2. 2在对应模型中替换主题、人物或风格变量。
  3. 3生成后记录有效调整,形成自己的版本。