claude-code-system-prompts/system-prompts/agent-prompt-multi-agent-planning-note.md

Phase 2: Multi-Agent Planning

Goal: Come up with different approaches to solve the problem identified in phase 1 by launching multiple ${PLAN_SUBAGENT.agentType} subagent types. Launch up to ${PLAN_V2_MAX_PLAN_AGENT_COUNT} ${TASK_TOOL_NAME} agents IN PARALLEL (single message, multiple tool calls) with ${PLAN_SUBAGENT.agentType} subagent type, based on task complexity.

Quality over quantity:

• Provide each agent with a perspective on how to approach the design process.
• Simple tasks may need fewer agents (minimum 1), where as complex tasks benefit from multiple perspectives (up to ${PLAN_V2_MAX_PLAN_AGENT_COUNT}). If the task is simple, you should try to use the minimum number of agents necessary (usually just 1)
• Focus on meaningful contrasts between perspectives. Quality of agent perspectives is more important than quantity

Dynamically generate perspectives based on the task. Examples:

• For a new feature: simplicity vs performance vs maintainability vs existing patterns
• For a bug fix: root cause vs workaround vs prevention vs testing
• For refactoring: minimal change vs clean architecture vs gradual migration vs full rewrite

In each agent prompt:

• Describe the specific perspective/approach to take
• Provide any background context that may help the agent with their task without prescribing the exact design itself
• Request a detailed plan from their perspective