Clawteam Openclaw Swarm

What This Skill Does

ClawTeam-OpenClaw-Swarm is a sophisticated orchestration layer designed to transform individual CLI-based coding agents—such as OpenClaw, Claude Code, or Codex—into a cohesive, self-organizing swarm. By leveraging git worktrees for isolated file environments and tmux for parallel process management, this skill enables a "leader" agent to delegate sub-tasks to multiple "worker" agents. It handles the complexity of task distribution, inter-agent messaging via a specialized inbox, and the final merging of code contributions. It essentially turns your terminal into a multi-agent control center, allowing you to tackle complex, large-scale coding projects that are too daunting for a single agent to manage alone.

Installation

To install, first ensure you have Python 3.10+ and tmux installed on your system. Because this is a specialized fork of the original ClawTeam project, you must clone the repository directly from the provided source: git clone https://github.com/win4r/ClawTeam-OpenClaw.git. Navigate into the folder and run pip install -e . for development mode.

For seamless integration, you must symlink the clawteam binary to ~/bin to ensure spawned agents can access the command within their fresh shells. If you are an OpenClaw user, you must also copy the skill definition file to your OpenClaw workspace and modify your security configuration to allow the agent to execute coordination commands without constant manual approval prompts. For convenience, the package includes an scripts/install-openclaw.sh file that automates the environment setup and configuration.

Use Cases

• Complex Refactoring: Assign one agent to modify database schemas while another updates the API layer, and a third writes the necessary unit tests.
• Feature Expansion: Spawn a swarm to implement a new feature set; one agent handles the backend logic, another handles the UI components, and the leader aggregates the results.
• Parallel Debugging: Assign different modules to separate agents to investigate and patch bugs simultaneously, significantly reducing incident resolution time.

Example Prompts

1. "@clawteam, spawn a swarm to build a new authentication module: Agent A handles the SQLite schema, Agent B writes the login endpoints, and Agent C documents the changes."
2. "@clawteam, coordinate the team to audit the current codebase for security vulnerabilities and prepare a summary report of findings."
3. "@clawteam, assign Agent D to fix the dependency conflicts in our project and report back to the main workspace once merged."

Tips & Limitations

• System Resources: Spawning multiple agents involves creating multiple git worktrees and tmux panes. Ensure your machine has sufficient CPU and RAM, as running too many simultaneous agents can lead to system instability.
• Security: Since this skill orchestrates agent behavior and file manipulation, always review the code generated by your swarms before committing to your main branch.
• Git Management: Because each worker uses its own git worktree, ensure your project is initialized as a git repository; otherwise, the swarm will fail to initialize agent environments.