Back to Details

multi-agent-workflow

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Multi-Agent Workflow Guide

多Agent工作流指南

When to use

适用场景

  • Complex feature spanning multiple domains (full-stack, mobile)
  • Coordination needed between frontend, backend, mobile, and QA
  • User wants step-by-step guidance for multi-agent coordination
  • 跨多个领域的复杂功能（全栈、移动端）
  • 需要协调前端、后端、移动端和QA Agent的场景
  • 用户需要多Agent协调的分步指导

When NOT to use

不适用场景

  • Simple single-domain task -> use the specific agent directly
  • User wants automated execution -> use orchestrator
  • Quick bug fixes or minor changes
  • 简单的单领域任务 -> 直接使用对应Agent即可
  • 用户需要自动执行 -> 使用orchestrator
  • 快速bug修复或小改动

Core Rules

核心规则

  1. Always start with PM Agent for task decomposition
  2. Spawn independent tasks in parallel (same priority tier)
  3. Define API contracts before frontend/mobile tasks
  4. QA review is always the final step
  5. Assign separate workspaces to avoid file conflicts
  6. Always use Serena MCP tools as the primary method for code exploration and modification
  7. Never skip steps in the workflow — follow each step sequentially without omission
  1. 始终从PM Agent开始进行任务分解
  2. 并行生成独立任务（同一优先级层级）
  3. 在前端/移动端任务前定义API契约
  4. QA审核始终是最后一步
  5. 分配独立工作区以避免文件冲突
  6. 始终使用Serena MCP工具作为代码探索和修改的主要方式
  7. 绝不跳过工作流中的步骤——按顺序执行每一步，不得遗漏

Workflow

工作流

Step 1: Plan with PM Agent

步骤1：与PM Agent规划

PM Agent analyzes requirements, selects tech stack, creates task breakdown with priorities.
PM Agent分析需求、选择技术栈、创建带优先级的任务分解。

Step 2: Spawn Agents by Priority

步骤2：按优先级生成Agent

Spawn agents via CLI:
  1. Use spawn-agent.sh for each task
  2. CLI selection follows agent_cli_mapping in user-preferences.yaml
  3. Spawn all same-priority tasks in parallel using background processes
bash
undefined
通过CLI生成Agent：
  1. 为每个任务使用spawn-agent.sh
  2. CLI选择遵循user-preferences.yaml中的agent_cli_mapping
  3. 使用后台进程并行生成所有同优先级任务
bash
undefined

Example: spawn backend and frontend in parallel

Example: spawn backend and frontend in parallel

oh-my-ag agent:spawn backend "task description" session-id -w ./backend & oh-my-ag agent:spawn frontend "task description" session-id -w ./frontend & wait
undefined
oh-my-ag agent:spawn backend "task description" session-id -w ./backend & oh-my-ag agent:spawn frontend "task description" session-id -w ./frontend & wait
undefined

Step 3: Monitor & Coordinate

步骤3：监控与协调

  • Use memory read tool to poll 
    progress-{agent}.md
    files
  • Verify API contracts align between agents
  • Ensure shared data models are consistent
  • 使用内存读取工具轮询
    progress-{agent}.md
    文件
  • 验证Agent之间的API契约是否一致
  • 确保共享数据模型保持统一

Step 4: QA Review

步骤4：QA审核

Spawn QA Agent last to review all deliverables. Address CRITICAL issues by re-spawning agents.
最后生成QA Agent以审核所有交付成果。通过重新生成Agent来解决关键问题。

Automated Alternative

自动化替代方案

For fully automated execution without manual spawning, use the orchestrator skill instead.
如果无需手动生成Agent，想要完全自动执行，请使用orchestrator技能。

References

参考资料

  • Workflow examples: 
    resources/examples.md
  • 工作流示例：
    resources/examples.md