calling-agent-squad

What This Skill Does

The calling-agent-squad skill is a sophisticated orchestrator designed to deploy multi-agent workflows within the OpenClaw environment. It transforms the agent from a single assistant into a specialized team of experts—including a Manager, Architect, Coder, Researcher, and Reviewer. This skill handles complex, multi-stage projects by managing the persona adoption and task distribution, ensuring that every project component receives specialized attention. It operates in two modes: Standard mode, where the main agent cycles through personas internally to maintain a consistent execution thread, and Full mode, which spawns independent agent instances with their own workspaces for maximum isolation and specialized compute power.

Installation

To integrate this capability into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/arbiger/calling-agent-squad

Ensure you have initialized the necessary agent personas by running the provided initialization script located at ~/.openclaw/workspace/skills/calling-agent-squad/squad-init.sh to populate the mandatory SOUL.md and IDENTITY.md configuration files.

Use Cases

This skill is ideal for projects requiring structured, high-quality output. It excels at:

• Software Engineering: Designing architecture, writing code, and conducting automated code reviews.
• Content Production: Researching topics, drafting copy, and performing editorial reviews to ensure brand alignment.
• Complex Problem Solving: Using the 'Observer' persona to audit workflows or analyze business strategy.

Example Prompts

1. "calling squad website-redesign Create a modern portfolio site using React and Tailwind CSS."
2. "calling squad full market-research Analyze the Q3 trends in the EV sector and draft a 5-page report."
3. "calling squad software-patch Document the security vulnerabilities in the current codebase and suggest remediation steps."

Tips & Limitations

• Standard Mode Efficiency: Use this for quick, iterative tasks where memory persistence isn't required across deep sub-agent processes.
• Full Mode Performance: This requires more tokens and processing power as it spawns isolated environments; reserve it for large-scale production projects.
• Data Integrity: Projects are automatically timestamped and saved in Documents/squad_projects/. Regularly clear old directories to prevent disk clutter.
• Security: Always verify the content of your SOUL.md files, as these define the constraints and behavioral boundaries for each squad member.