openclaw-multi-agent

Coordinate hub-and-spoke OpenClaw multi-agent systems with session-based routing and sandbox controls.

Updated Mar 21, 2026
One-click install
npx skills add https://github.com/kyle-deprow/ai_scaffolding --skill openclaw-multi-agent-kyle-deprow
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: openclaw-multi-agent
Source: https://github.com/kyle-deprow/ai_scaffolding/tree/main/skills/openclaw-multi-agent
Command: npx skills add https://github.com/kyle-deprow/ai_scaffolding --skill openclaw-multi-agent-kyle-deprow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

OpenClaw multi-agent orchestration tackles the challenge of coordinating multiple specialized agents, enabling session-based communication, controlled tool access, and structured delegation patterns in complex workflows.

Core Features & Use Cases

  • Spawning subagents for delegated tasks and managing their lifecycles.
  • Inter-agent messaging and session-based routing across channels.
  • Sandbox visibility controls, agent allowlists, and per-spawn tool scoping for secure orchestration.
  • Use Case: design scalable workflows where a coordinator delegates research, coding, and summarization tasks and aggregates results.

Quick Start

Deploy a Claw controller to spawn a researcher, configure basic routing, and initiate a test inter-agent session.

Frequently Asked Questions about openclaw-multi-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate a multi-agent system for complex task delegation?

Multi-agent orchestration coordinates a hub-and-spoke system to delegate subagents, sessions, and routing. A defined coordinator agent spawns specialized subagents for research, coding, and summarization while aggregating results.

What is session-based routing in multi-agent workflows?

Session-based routing enables inter-agent messaging across channels using tools like sessions_list and sessions_send. It allows a coordinator to send tasks to spawned subagents and receive structured responses.

How do I scope tool access when spawning subagents?

Per-spawn tool scoping restricts subagent capabilities through agent allowlists and per-spawn tool subsets. This ensures spawned researchers, coders, and writers only access authorized tools.

Do I need a coordinator agent to manage multi-agent sessions?

Yes, a defined coordinator agent like claw is required to manage subagent lifecycles, configure routing, and initiate inter-agent sessions. It controls sandbox visibility, agent allowlists, and tool subsets.

What's the best way to securely manage subagent lifecycles in OpenClaw?

Secure subagent lifecycle management uses sandbox visibility controls and agent allowlists. The coordinator spawns subagents with restricted tool subsets, delegates tasks, and aggregates results safely.

Can I use sandbox visibility controls to restrict agent access in multi-agent orchestration?

Yes, sandbox visibility controls restrict agent access during orchestration. Combined with agent allowlists and per-spawn tool subsets, they secure delegation patterns across complex workflows.