naruto-multi-agent-cn

What This Skill Does

The naruto-multi-agent-cn skill transforms your OpenClaw agent into the Fifth Hokage, Tsunade. It implements a sophisticated multi-agent orchestration architecture designed to delegate work to five specialized persistent sub-agents (Naruto, Kakashi, Shikamaru, Sakura, and Sai). By utilizing session reuse through sessions_spawn and a round-robin scheduling protocol, the skill ensures that long-running tasks maintain contextual continuity. The agent acts strictly as a coordinator, enforcing a 'speak-before-spawn' protocol that forces the agent to communicate its assessment of the task (ranked S to D) to the user before triggering any sub-agent actions. It also supports intelligent multi-task decomposition, allowing the agent to parse complex user requests and dispatch multiple sub-agents in parallel when tasks are independent.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/be1human/naruto-multi-agent-cn

Use Cases

This skill is ideal for complex development workflows where you need a 'Project Manager' persona to handle technical tasks. It is perfect for:

1. Refactoring Projects: Dispatching the 'Kakashi' agent for deep architecture analysis and 'Sakura' for debugging.
2. Large Feature Development: Using the 'Naruto' agent for heavy coding and 'Shikamaru' for strategy and logic planning.
3. Research & Documentation: Utilizing 'Sai' for extracting information and documentation drafting across multiple threads.
4. Task Prioritization: Ensuring tasks are categorized by danger levels (S-D) to prioritize effort effectively.

Example Prompts

1. "I need to refactor the authentication module and update the README.md file. Can you handle this?"
2. "Perform a security audit on the current API structure, and then compile a list of all potential performance bottlenecks."
3. "Analyze the current frontend state management and propose a plan to migrate to a global store."

Tips & Limitations

• Roleplay Adherence: The agent is designed to stay in character. Use natural language to get the best responses.
• Session Management: The skill relies heavily on the sessionKey parameter. Avoid manually creating sessions; rely on the skill's automated dispatching to ensure the persistent agents retain their memory.
• Constraint: The primary agent cannot perform execution tasks itself. If you attempt to force the coordinator to write files directly, it will prompt you to delegate the work to a sub-agent. Ensure your requests are clear to allow the coordinator to accurately classify them as S, A, B, C, or D level missions.