war-room

What This Skill Does

The War Room skill is a high-level orchestration framework designed for OpenClaw to conduct structured, multi-agent brainstorming and project development. It shifts the AI workflow from single-agent linear tasks to a collaborative environment where specialized agents act as stakeholders, engineers, and critics. By utilizing a shared filesystem and a wave-based dependency protocol, the skill ensures that complex problems—ranging from software architecture to business strategy—are vetted by a 'CHAOS' agent that acts as a perpetual devil's advocate. This approach prevents common pitfalls by forcing teams to confront the 'riskiest assumption' in Wave 0, ensuring that time and tokens are not wasted on fundamentally flawed projects.

Installation

To integrate this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/maxkle1nz/war-room Ensure your current environment has write permissions to the workspace directory to facilitate the creation of the war-rooms/ folder structure.

Use Cases

• System Architecture: Design complex cloud infrastructure or software systems by having networking, database, and security agents collaborate simultaneously.
• Business Strategy: Stress-test a new product launch plan, incorporating roles like 'Chief Economist', 'Marketing Lead', and 'Regulatory Counsel'.
• Product Specs: Develop high-fidelity documentation by iterating through project requirements with specialist feedback loops.
• Troubleshooting: Use the War Room to perform deep-dive post-mortems or architectural analysis on struggling systems, using the CHAOS agent to uncover hidden performance bottlenecks.

Example Prompts

1. "OpenClaw, start a new war room for a high-frequency trading bot project. Initialize the folder structure and let's brainstorm the core architecture."
2. "I have an idea for a decentralized social media app. Let's run a Wave 0 test. What is the single riskiest assumption we need to prove before building?"
3. "Run a war room session on our current project and have the CHAOS agent challenge our latest API integration decisions in the architecture doc."

Tips & Limitations

• Dependency Tracking: Always define your dependency graph clearly. If Wave 2 agents require output from Wave 1, ensure the directory structure is updated correctly by the agents.
• Token Management: Because this skill spawns multiple subagents, be mindful of token consumption. Keep individual waves concise and focused.
• The Pivot Gate: Do not skip the pivot gate. It is the most critical feature for preventing 'hallucinated momentum' where agents build on invalid logic.
• System Integrity: The skill relies heavily on the BRIEF.md and DNA.md files; keep these updated to ensure agents maintain consistent context throughout the session.