← Back to Skills Marketplace
91
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install agent-soul-system
Description
Create, manage, and validate SOUL.md personality files for multi-agent systems. Provides a standardized soul architecture where each agent has a distinct per...
README (SKILL.md)
Agent Soul System
Standardized personality system for multi-agent architectures. Each agent has a distinct SOUL.md defining personality, principles, and collaboration protocol.
Directory Structure
~/.openclaw/agents/
├── [agent-name]/
│ ├── SOUL.md ← 灵魂（人格+原则+职责）
│ ├── AGENTS.md ← 工作规则
│ ├── IDENTITY.md ← 身份卡
│ └── MEMORY.md ← 长期记忆
SOUL.md Standard Structure
Every agent's SOUL.md follows this template:
# SOUL - [Agent名字]
## 人格
[历史/现实人物] — [核心理念]
## 核心特质
- 从导师继承的特质1
- 从导师继承的特质2
## 说话风格
- 典型口头禅/句式
## 核心职责
1. 职责1
2. 职责2
## 核心原则
1. 原则1
2. 原则2
## 与上级的协作协议
- 如何汇报
- 什么级别需要上报
## 输出标准
- 产出格式要求
Quick Commands
# Validate all agents have SOUL.md
python3 ~/.openclaw/workspace/skills/agent-soul-system/scripts/soul-check.py
# Create a new agent with SOUL.md template
python3 ~/.openclaw/workspace/skills/agent-soul-system/scripts/soul-create.py [agent-name]
# List all agents and their souls
python3 ~/.openclaw/workspace/skills/agent-soul-system/scripts/soul-ls.py
Personality Library
See references/personality-library.md for a curated collection of historical/contemporary figures with their core traits, suitable for agent persona assignment.
Reference Files
| File | Purpose |
|---|---|
references/soul-template.md |
Standard SOUL.md template |
references/personality-library.md |
Figure-based persona catalog |
references/collaboration-protocol.md |
Multi-agent collaboration patterns |
Workflow
New Agent Setup
- Run
soul-create.py [name]to generate SOUL.md skeleton - Select personality from
references/personality-library.md - Fill in职责, 原则, 协作协议
- Run
soul-check.pyto validate
Audit
- Run
soul-check.pyto scan all agents - Review missing or malformed SOUL.md files
- Fix issues and re-validate
Collaboration Setup
- Define hierarchy (who reports to whom)
- Set escalation rules in each SOUL.md
- Run
soul-check.py --collabto verify protocol consistency
Current Agent Roster (Example)
| Agent | Personality | Role |
|---|---|---|
| 小咪 | 直接高效+幽默 | 指挥官、系统大脑 |
| Jarvis Invest | 毛泽东+索罗斯 | 投资交易 |
| sains-crm | 张良 | 商机管理 |
| personal-cfo | 查理·芒格 | 财务分析 |
| 休眠修炼 | 诸葛亮+德鲁克 | 技术进化 |
| family-guardian | 孔子+曾国藩 | 家庭守护 |
| agent-army | 韩信 | 多Agent调度 |
| main | 福尔摩斯 | 通用备用 |
Usage Guidance
What to consider before installing: (1) This skill is local-only and reads/writes files under ~/.openclaw/agents and uses bundled reference files; it does not use the network or request credentials. (2) If you already have agents under ~/.openclaw/agents, back up that directory first — soul-create.py can create or overwrite SOUL.md (it prompts before overwriting). (3) Review the three included Python scripts (soul-create.py, soul-ls.py, soul-check.py) yourself if you want to confirm behavior; they're short and only perform filesystem operations and stdout printing. (4) Be aware the personality library includes historical and political figures — that is content, not functionality; decide whether those examples are acceptable for your environment. (5) If you prefer additional safety, run the scripts in a sandbox or a test account before using them on production agent data.
Capability Analysis
Type: OpenClaw Skill
Name: agent-soul-system
Version: 1.0.0
The agent-soul-system bundle is a framework for managing agent personas and collaboration protocols. The included Python scripts (soul-check.py, soul-create.py, and soul-ls.py) perform standard file I/O operations within the ~/.openclaw/agents directory to create and validate markdown-based personality files. There is no evidence of data exfiltration, network activity, or malicious execution; the bundle functions exactly as described in its documentation.
Capability Assessment
Purpose & Capability
Name/description align with the included scripts and reference files. The three Python scripts implement creation, listing, and validation of SOUL.md files and the references match the stated persona/collaboration features.
Instruction Scope
The SKILL.md and scripts operate on the user's filesystem (AGENTS_DIR = ~/.openclaw/agents) and on files bundled in the skill workspace (references/*). Metadata declared no required config paths, but the runtime instructions clearly read and write under ~/.openclaw/agents — this is expected for the skill's purpose but is worth noting because it accesses user files.
Install Mechanism
No install spec or external downloads. All code is bundled with the skill (three small Python scripts and markdown refs). No network fetches or archive extraction observed.
Credentials
The skill requires no environment variables, no credentials, and the scripts do not read environment secrets. That matches the local, filesystem-only functionality.
Persistence & Privilege
The skill is not always-enabled and does not request elevated privileges, but it will create directories and write SOUL.md files under ~/.openclaw/agents. soul-create.py will prompt before overwriting existing SOUL.md, but the skill does persist data to the user's home directory.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install agent-soul-system - After installation, invoke the skill by name or use
/agent-soul-system - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of agent-soul-system: a standardized personality and collaboration framework for multi-agent systems.
- Provides tools to create, manage, and validate SOUL.md files defining agent personalities.
- Includes directory and file structure conventions for agent setup and long-term memory.
- Supplies a SOUL.md template covering personality, traits, collaboration protocols, and output standards.
- Features quick commands for agent creation, auditing, and listing via command-line scripts.
- References a curated personality library for assigning historical or philosophical personas to agents.
- Supports personality-driven team building and protocol validation for effective multi-agent collaboration.
Metadata
Frequently Asked Questions
What is Agent Soul System?
Create, manage, and validate SOUL.md personality files for multi-agent systems. Provides a standardized soul architecture where each agent has a distinct per... It is an AI Agent Skill for Claude Code / OpenClaw, with 91 downloads so far.
How do I install Agent Soul System?
Run "/install agent-soul-system" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Agent Soul System free?
Yes, Agent Soul System is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Agent Soul System support?
Agent Soul System is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Agent Soul System?
It is built and maintained by Adam (@adamwgp); the current version is v1.0.0.
More Skills