openclaw-agent-development

Configure autonomous agents with IDENTITY.md, SOUL.md, and openclaw.json.

4|Updated Feb 26, 2026
One-click install
npx skills add https://github.com/oabdelmaksoud/Openclaw-skills-Compilations --skill openclaw-agent-development
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: openclaw-agent-development
Source: https://github.com/oabdelmaksoud/Openclaw-skills-Compilations/tree/main/openclaw-agent-development
Command: npx skills add https://github.com/oabdelmaksoud/Openclaw-skills-Compilations --skill openclaw-agent-development

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

OpenClaw agent development provides a structured blueprint for creating, configuring, and managing autonomous agents in a 16-agent OpenClaw system. It explains how to compose per-agent identities (IDENTITY.md) and shared instructions (SOUL.md), how to model routing in openclaw.json, and how to organize agent workspaces for consistent collaboration.

Core Features & Use Cases

  • Guidance to build per-agent workspaces at ~/.openclaw/workspace/agents-workspaces/<id>/
  • Templates for IDENTITY.md and SOUL.md with best practices
  • JSON configuration patterns for agents and model routing
  • On-demand patterns for comms, authentication, and tool usage
  • Examples of full agent configurations and spawn flows

Quick Start

Create a new agent workspace, write IDENTITY.md and SOUL.md templates, and register the agent in openclaw.json.

Frequently Asked Questions about openclaw-agent-development

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

FAQPage Schema
How do I configure autonomous agent identity and shared behavior in a multi-agent workspace?

To configure autonomous agents, define per-agent identity in IDENTITY.md and shared instructions in SOUL.md, then register these templates within the agent workspace directory to establish consistent collaboration behavior.

What is the best way to route models for autonomous agents in a multi-agent system?

The best way to route models for autonomous agents is by defining model routing patterns and agent configurations in a central openclaw.json file, ensuring smooth orchestration and communication across the 16-agent system.

How do I set up agent workspaces with inbox and outbox channels for orchestration?

Set up agent workspaces by creating directories at ~/.openclaw/workspace/agents-workspaces/<id>/ and configuring inbox and outbox channels to enable seamless message passing and orchestration between agents.

Do I need separate identity files for each agent in a 16-agent system?

Yes, a 16-agent system requires a separate IDENTITY.md file for each agent to define its specific role, while a shared SOUL.md file provides the common behavioral instructions across all autonomous agents.

What is the difference between IDENTITY.md and SOUL.md when designing autonomous agents?

IDENTITY.md defines the unique per-agent configuration and role, whereas SOUL.md establishes the shared instructions and common behavior patterns applied uniformly across all agents in the workspace.

Can I use JSON configuration patterns to manage agent spawn flows and authentication?

Yes, JSON configuration patterns in openclaw.json support on-demand mechanisms for comms, authentication, tool usage, and defining full agent configurations including spawn flows.