This project provides a streamlined, one-command setup for a multi-agent system tailored for OpenClaw, enabling researchers and power users to rapidly deploy a fleet of specialized AI agents. It simplifies the complex task of configuring collaborative AI teams for tasks like research paper production, brainstorming, and analysis, significantly reducing setup time and effort.

How It Works

OpenClaw Agents provisions nine pre-configured, specialized AI agents designed for collaborative research workflows. The system emphasizes adversarial collaboration, featuring dynamics like Ideator vs. Critic and Writer vs. Reviewer, to foster creative tension and rigorous quality control. Key design elements include "SHARP taste gates" enforced by a Critic agent and a "safe merge" mechanism that integrates new agents without overwriting existing OpenClaw configurations.

Quick Start & Requirements

Installation is primarily handled via a bash script (setup.sh) after cloning the repository. Prerequisites include the OpenClaw CLI (installable via npm install -g openclaw@latest) and standard shell utilities. The setup script interactively guides users through selecting deployment modes (Channel Mode for chat platforms like Feishu, WhatsApp, Telegram, Discord, or Local Workflow Mode), specifying channels, assigning group IDs, and configuring LLM models (defaulting to zai/glm-5, with options for per-agent customization via --model-map).

Highlighted Details

• Nine Specialized Agents: A core fleet including Planner, Ideator, Critic, Surveyor, Coder, Writer, Reviewer, Scout, and a Main orchestrator, each with distinct roles and emoji identities.
• Workflow Templates: Four pre-built workflows are provided: Paper Pipeline, Daily Digest, Brainstorm, and Rebuttal.
• Adversarial Collaboration: Built-in productive tension between agents like Ideator↔Critic and Writer↔Reviewer.
• SHARP Taste Gates: Quality checkpoints enforced by the Critic agent at critical stages of the research process.
• Safe Merge: Ensures agents are appended to existing openclaw.json configurations, preserving user settings.
• Flexible Deployment: Supports direct integration with popular chat platforms or local communication via the agentToAgent tool.

Maintenance & Community

The project is maintained by its creator, with contributions welcomed via pull requests. Specific community channels (e.g., Discord, Slack) or detailed contributor information are not explicitly detailed in the provided README.

Licensing & Compatibility

The project is released under the permissive MIT License, allowing for free use, modification, and distribution, including for commercial purposes and integration with closed-source projects.

Limitations & Caveats

The eight core agent IDs (e.g., planner, ideator) are protected and should not be modified. Agents operate within independent workspaces rather than a Docker sandbox. While the setup script simplifies deployment, a foundational understanding of OpenClaw and LLM configurations is beneficial. The system relies on specific LLM models, necessitating access or alternative configurations.