A sophisticated multi-agent orchestration system that organizes AI models into specialized professional teams including software development, business strategy, and creative writing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install team-collab
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install team-collab using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Team Collab: Multi-Model Multi-Agent Collaboration System?
Team Collab is a high-performance framework designed for Openclaw Skills that transforms standalone AI models into cohesive, professional business units. By utilizing a fixed sub-window architecture, it assigns specific roles to the most capable models—such as qwen3.5-plus for creativity and glm-5 for logic—ensuring each task is handled by an expert agent with persistent memory.
This skill enables users to operate like a full-scale company, providing structured workflows for complex projects. Whether you are building software, launching a commercial venture, or drafting a web novel, Team Collab coordinates up to 28 specialized AI roles to deliver professional-grade results through a unified collaboration protocol.
Team Collab: Multi-Model Multi-Agent Collaboration System Use Cases
- Orchestrating a full software development lifecycle from requirements to code review.
- Launching and managing commercial monetization projects with the specialized money-making squad.
- Professional AI novel creation and manga scriptwriting using the specialized fiction team.
- Managing complex multi-step technical tasks that require distinct expert perspectives.
- Maintaining persistent, role-based AI interactions through dedicated sub-window sessions.
How Team Collab: Multi-Model Multi-Agent Collaboration System Works
- The user triggers a specific team using keywords like 软件开发团队 (Software Team), 搞钱特战队 (Business Team), or 小说团队 (Novel Team).
- The system spawns or activates persistent sub-windows for each predefined role within that team.
- Each sub-window is automatically assigned its optimal LLM (e.g., glm-5 for coding, qwen3.5-plus for creative content).
- Agents follow a structured cross-functional workflow, passing outputs from one expert (like a World Architect) to the next (like a Plot Engineer).
- Quality control agents and legal advisors review the final output to ensure technical accuracy and platform compliance.
Team Collab: Multi-Model Multi-Agent Collaboration System Setup
To integrate this skill into your environment, follow the standard Openclaw Skills installation process:
# Install the team-collab skill
openclaw install team-collab
# Configure model API keys for qwen3.5-plus and glm-5 in your .env
# Start the multi-agent environment
openclaw run team-collab
Team Collab: Multi-Model Multi-Agent Collaboration System Data Schema & Taxonomy
Team Collab uses a standardized naming convention for its sub-window architecture to maintain state and organizational clarity:
| Team Prefix | Role Key | Persistent Label |
|---|---|---|
| dev- | pm, coder, test, review | dev-[role] |
| money- | hunter, advisor, tech | money-[role] |
| novel- | expert, world, plot | novel-[role] |
All session data and generated artifacts are indexed by these labels, allowing for multi-agent memory persistence across different conversation segments.
Team Collab: Multi-Model Multi-Agent Collaboration System Advanced Features
- Fixed Sub-Window Mode: Ensures each role has a dedicated, persistent workspace for long-term project memory.
- Optimal Model Routing: Automatically assigns specific tasks to the LLM best suited for the job (Creative vs. Logical).
- Integrated Quality Assurance: Dedicated 'Quality Control' agents in every team specifically for bug fixing and output validation.
- Multi-Team Cross-Pollination: Ability to trigger multiple professional squads within the same Openclaw Skills environment.
SKILL.md
Loading
Related Openclaw Skills
Feishu ASR (Whisper)
A privacy-first speech recognition skill that transcribes Feishu voice messages locally using OpenAI Whisper models.
Poof
A versatile integration skill for managing ephemeral messaging data, deals, leads, and project activities through the Membrane ecosystem.
chen-Image
A comprehensive image processing and optimization skill designed to handle professional workflows for web delivery, social media, ecommerce, and print.
Book Processor
An automated utility for transforming EPUB files into structured knowledge assets including summaries, checklists, and thinking models.
Blast L2 — Ethereum L2 Chain Reference
A comprehensive technical reference skill for navigating the Blast L2 ecosystem, focusing on native yield mechanics and developer infrastructure.
Baidu Scholar Search
A specialized research skill for searching Chinese and English academic literature, journals, and dissertations via Baidu Scholar.
Featured*
