A master-worker agent cluster framework designed for parallel task execution and distributed AI workflows.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-weave
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 agent-weave 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 Agent-Weave?
Agent-Weave is a robust orchestration layer designed for managing complex agent architectures and distributed systems. It allows developers to build scalable environments where a master agent coordinates multiple worker agents to handle high-concurrency tasks. By integrating this into your library of Openclaw Skills, you can implement sophisticated MapReduce-style workflows and ensure secure, isolated communication between parent and child processes. This skill provides the necessary infrastructure for anyone looking to scale their agentic task execution beyond the limitations of single-threaded or single-agent environments.
Agent-Weave Use Cases
- Building distributed agent systems that require high-throughput parallel processing.
- Orchestrating multiple specialized agents to collaborate on complex, multi-step goals.
- Implementing MapReduce-style workflows for data-intensive AI operations.
- Scaling task execution dynamically across a fleet of managed worker agents.
- Creating resilient master-worker architectures for enterprise-grade automation within Openclaw Skills.
How Agent-Weave Works
- Initialize the Loom factory which serves as the primary manager for the lifecycle of your agent cluster.
- Create a Master agent that acts as the central orchestrator and task dispatcher for the cluster.
- Spawn multiple Worker agents that are securely bound to the parent Master through a dedicated communication layer.
- Dispatch data or tasks from the Master agent, which automatically distributes the load across all active Workers.
- Workers process their assigned tasks in parallel and return the processed data to the Master.
- The Tapestry engine aggregates the final results, allowing for complex data reduction and orchestration logic.
Agent-Weave Setup
To begin using this tool within your Openclaw Skills environment, install the package via npm:
npm install agent-weave
You can also manage your clusters directly through the CLI:
# Create a new master agent named my-cluster
weave loom create-master --name my-cluster
# Spawn 5 worker agents for the specified master
weave loom spawn --parent <master-id> --count 5
# Visualize the agent hierarchy
weave loom list --tree
Agent-Weave Data Schema & Taxonomy
Agent-Weave organizes its internal data and cluster relationships through a hierarchical metadata taxonomy to ensure reliable task tracking.
| Component | Role | Description |
|---|---|---|
| Loom | Factory | The root object that manages agent creation and global state tracking. |
| Master | Orchestrator | Stores cluster IDs, worker registry, and the active task queue. |
| Worker | Executor | Independent nodes responsible for processing individual data packets assigned by the Master. |
| Thread | Communication | The secure channel that handles parent-child event emitting and data transfer. |
| Tapestry | Orchestrator | The engine that defines the logical flow for MapReduce and task distribution. |
Agent-Weave Advanced Features
- Dynamic Auto-scaling: Automatically manage worker counts based on real-time task load and system performance.
- Secure Threading: Enforced parent-child communication protocols to ensure data integrity across Openclaw Skills.
- MapReduce Support: Built-in Tapestry engine for sophisticated data splitting and aggregation workflows.
- Event-driven Monitoring: Fully compatible with EventEmitter for real-time status updates and cluster health tracking.
- Hierarchical Visualization: Built-in CLI tools to inspect and debug complex agent trees and communication paths.
SKILL.md
Loading
Related Openclaw Skills
Voice.ai Voice Agents
A comprehensive tool for building, managing, and deploying professional conversational voice agents via the Voice.ai Agent API.
Dub YouTube with Voice.ai
An automated workflow for generating studio-quality AI voiceovers, YouTube chapters, and SRT captions from text scripts using Voice.ai.
QA Check
A comprehensive quality assurance suite designed to validate web projects and builds before they go live to production.
GizmoLab Tools & UI
A comprehensive suite of blockchain developer utilities and React-based Web3 UI components for Ethereum and Solana.
index1 Doctor
A specialized diagnostic utility that verifies the technical health, dependencies, and model configuration of index1 environments.
index1
A persistent AI memory system providing coding agents with hybrid search, code indexing, and cognitive fact recording across sessions.
Featured*
