Setup & Installation
openclaw skills install @gl813788-byte/agent-weave
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install agent-weave
What This Skill Does
agent-weave is a Node.js library for building master-worker agent clusters with parallel task execution. It provides a Loom factory, Master orchestrator, and Worker nodes that communicate through a secure parent-child channel. Supports MapReduce-style workflows via a built-in Tapestry engine.
Handles the boilerplate of inter-agent communication and worker lifecycle so you don't have to wire up message passing and task dispatch manually.
When to use it
- Splitting a large dataset across 10 parallel worker agents
- Running concurrent web scraping jobs with a single coordinator
- Distributing image processing tasks to dynamically scaled workers
- Implementing a map-reduce pipeline for log aggregation
- Orchestrating multi-step agent workflows with dependency ordering
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User asks: process an array of 1000 numbers in parallel using 5 workers
AGENT
- 1Import Loom from agent-weave and instantiate it
- 2Call loom.createMaster('my-cluster') to create a master node
- 3Call loom.spawnWorkers(master.id, 5, handler) to create 5 workers with a processing function
- 4Call master.dispatch(dataArray) to distribute tasks across workers
- 5Collect and return the aggregated results array
OUTPUT
Array of 1000 processed results, each handled by one of the 5 parallel workers