A sophisticated AI orchestration engine that builds, manages, and evaluates specialized agent teams for complex engineering and development workflows.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install teamwork
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 teamwork 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 Teamwork?
The Teamwork skill is a high-level orchestration framework designed to manage multi-agent collaboration for projects that exceed the capacity of a single AI model. By utilizing Openclaw Skills, this system allows users to decompose complex requirements into specialized roles like Architects, Developers, and Testers. It ensures that every task is matched with the most suitable model based on real-time performance metrics and budget constraints.
This skill introduces a structured hierarchy involving a Host Model for user interaction and a Herald model for inter-agent communication. It maintains a continuous feedback loop through peer-evaluation, ensuring that your AI team becomes more efficient over time by tracking model performance across eight critical dimensions, including thinking depth, code quality, and reliability.
Teamwork Use Cases
- Executing large-scale software development projects requiring specialized analysis, coding, and QA roles.
- Balancing high-performance reasoning tasks with cost-effective models to stay within monthly budgets.
- Automated system design where multiple models must reach a consensus on architectural patterns.
- Continuous integration and code review workflows where models provide peer-to-peer feedback.
How Teamwork Works
- The Host Model analyzes the user's request and decomposes it into a detailed task tree with specific capability requirements.
- A team assembly meeting is convened where available models review the briefing and collaboratively define necessary roles.
- Roles are assigned to models using a weighted algorithm that considers capability scores, cost-efficiency, and current workload.
- A Herald model is designated to act as the central communication hub, relaying messages and monitoring task heartbeats.
- Agents execute tasks in parallel, reporting progress to the Herald who escalates any bottlenecks or failures to the Host.
- After completion, a review meeting is held where models perform peer evaluations to update the global model scores database.
Teamwork Setup
The Teamwork skill features an autonomous initialization system. When first invoked, it will guide you through an interactive CLI setup to configure your environment.
- Run the initialization script to create the necessary directory structure:
mkdir -p .trae/config .trae/data .trae/skills/teamwork
- Provide API keys and base URLs for your preferred AI providers (e.g., OpenAI, Anthropic).
- Define model pricing (Subscription, Tiered, or Pay-Per-Use) to enable the cost-optimization engine.
- Set your Host Model and monthly budget limits to prevent overages.
Teamwork Data Schema & Taxonomy
The skill manages its state and intelligence through a structured JSON schema stored in the project's metadata directory:
| Component | File Path | Description |
|---|---|---|
| Provider Config | .trae/config/providers.json |
Stores model capabilities, API endpoints, and pricing models. |
| Role Definitions | .trae/config/team-roles.json |
Maps engineering roles to preferred model traits and required skills. |
| Score Database | .trae/data/model_scores.json |
Maintains historical performance metrics and role-fit assessments for all models. |
Teamwork Advanced Features
- Peer-to-peer evaluation system that updates model capability scores after every successful task execution.
- Dynamic Herald selection prioritizing the fastest-responding models for communication overhead reduction.
- Support for complex pricing models including daily/monthly quotas and tiered usage overage rates.
- Automated failure recovery protocols that include root-cause analysis and team reconfiguration upon task timeout.
SKILL.md
Loading
Related Openclaw Skills
App Legal Pages
Automate the creation and deployment of professional Privacy Policy and Terms of Service static sites directly from your app feature documentation.
Mood Sticker
An intelligent emotion-sensing skill that enables AI agents to autonomously share relevant stickers and memes based on conversation context.
Receiving Code Review
A technical protocol for AI agents to process code review feedback through rigorous verification and evidence-based implementation rather than blind agreement.
Executing Plans
A systematic protocol for AI agents to execute complex implementation plans through batch processing and mandatory review checkpoints.
OpenClaw Migration
A systematic migration playbook for transitioning codebases, documentation, and tooling from the Clawd brand to OpenClaw.
Invoice QR Scanner
A smart automation tool that extracts data from invoice QR codes and automatically completes electronic invoice applications.
Featured*
