Workspace Customization Examples
Complete end-to-end examples of OpenClaw workspace configurations for different industries and use cases. Each scenario shows all workspace files working together.
Available Scenarios
| Scenario |
Use Case |
Key Features |
| Legal Firm |
Law firm with legal research, compliance, client management |
Strict accuracy rules, citation requirements, confidentiality |
| Dev Team |
Software development team with CI/CD, code review, deployments |
Technical precision, deployment safety, code-first communication |
| Personal Assistant |
Individual productivity, scheduling, research, communications |
Proactive, warm tone, broad capability, privacy-focused |
| Marketing Agency |
Content creation, social media, client brand management |
Creative + analytical, brand voice matching, approval workflows |
| Customer Support |
Support team handling tickets, FAQs, escalations |
Empathetic, resolution-focused, escalation rules, SLA tracking |
| Content Creator |
Individual content creator, writing, research, publishing |
Creative partner, voice matching, workflow support |
How to Use These Scenarios
- Pick the closest scenario to your use case
- Read the full scenario from references/scenarios/
- Adapt — replace placeholders with your specific details
- Combine elements — take the SOUL.md from one scenario and AGENTS.md from another if your use case is hybrid
- Start minimal — don't copy everything; begin with SOUL.md + AGENTS.md + USER.md, add others as needed
Common Patterns Across All Scenarios
SOUL.md Pattern
Every good SOUL.md has:
- Domain-specific expertise statement
- 3-5 core values relevant to the domain
- Clear boundaries (what NOT to do)
- Communication style guidance
- Under 2,000 words
AGENTS.md Pattern
Every good AGENTS.md has:
- Session startup checklist
- Memory management rules
- Red lines / safety rules
- Group chat behavior (if applicable)
- Tool usage priorities
- Complex task handling rules
USER.md Pattern
Every good USER.md has:
- Primary user profile (name, timezone, language)
- Communication preferences
- Domain expertise level
- Approval authority
Workspace Structure Pattern
Every well-organized workspace has:
workspace/
├── Core files (AGENTS, SOUL, USER, IDENTITY, TOOLS)
├── HEARTBEAT.md (if periodic tasks needed)
├── MEMORY.md (curated long-term memory)
├── docs/
│ ├── procedures/ (domain-specific workflows)
│ ├── standing-orders/ (autonomous programs)
│ └── [domain]/ (domain-specific reference docs)
└── memory/ (daily logs, auto-managed)
Quick Start: Minimal Viable Workspace
For any use case, start with these three files:
1. SOUL.md (under 500 words to start)
## Core Truths
- [2-3 domain-specific values]
- Be resourceful before asking
## Boundaries
- [2-3 hard limits specific to your domain]
## Vibe
[One sentence describing communication style]
2. AGENTS.md (under 1,000 words to start)
## Session Startup
1. Read SOUL.md, USER.md
2. Read today's memory file
## Safety
- Don't run destructive commands without confirmation
- Keep private data private
## Memory
- Daily logs in memory/YYYY-MM-DD.md
- Long-term facts in MEMORY.md (main session only)
3. USER.md (minimal profile)
- Name: [your name]
- Timezone: [your timezone]
- Language: [preferred language]
- Style: [brief/detailed, formal/casual]
Then iterate: add IDENTITY.md, TOOLS.md, HEARTBEAT.md, and docs/ as needs arise.
1---2name: examples-393description: Complete end-to-end workspace customization examples for different industries — legal firms, dev teams, marketing agencies, customer support, personal assistants, and content creators. Use this skill whenever the user wants to see a reference implementation, asks "how would you set up OpenClaw for X", wants to see what a complete workspace looks like, or needs a starting template for their industry. Even vague requests like "show me an example" or "what does a good workspace look like" should trigger this skill.4---56# Workspace Customization Examples78Complete end-to-end examples of OpenClaw workspace configurations for different industries and use cases. Each scenario shows all workspace files working together.910## Available Scenarios1112| Scenario | Use Case | Key Features |13|----------|----------|-------------|14| [Legal Firm](references/scenarios/legal-firm.md) | Law firm with legal research, compliance, client management | Strict accuracy rules, citation requirements, confidentiality |15| [Dev Team](references/scenarios/dev-team.md) | Software development team with CI/CD, code review, deployments | Technical precision, deployment safety, code-first communication |16| [Personal Assistant](references/scenarios/personal-assistant.md) | Individual productivity, scheduling, research, communications | Proactive, warm tone, broad capability, privacy-focused |17| [Marketing Agency](references/scenarios/marketing-agency.md) | Content creation, social media, client brand management | Creative + analytical, brand voice matching, approval workflows |18| [Customer Support](references/scenarios/customer-support.md) | Support team handling tickets, FAQs, escalations | Empathetic, resolution-focused, escalation rules, SLA tracking |19| [Content Creator](references/scenarios/content-creator.md) | Individual content creator, writing, research, publishing | Creative partner, voice matching, workflow support |2021## How to Use These Scenarios22231. **Pick the closest scenario** to your use case242. **Read the full scenario** from references/scenarios/253. **Adapt** — replace placeholders with your specific details264. **Combine elements** — take the SOUL.md from one scenario and AGENTS.md from another if your use case is hybrid275. **Start minimal** — don't copy everything; begin with SOUL.md + AGENTS.md + USER.md, add others as needed2829## Common Patterns Across All Scenarios3031### SOUL.md Pattern32Every good SOUL.md has:33- Domain-specific expertise statement34- 3-5 core values relevant to the domain35- Clear boundaries (what NOT to do)36- Communication style guidance37- Under 2,000 words3839### AGENTS.md Pattern40Every good AGENTS.md has:41- Session startup checklist42- Memory management rules43- Red lines / safety rules44- Group chat behavior (if applicable)45- Tool usage priorities46- Complex task handling rules4748### USER.md Pattern49Every good USER.md has:50- Primary user profile (name, timezone, language)51- Communication preferences52- Domain expertise level53- Approval authority5455### Workspace Structure Pattern56Every well-organized workspace has:57```58workspace/59├── Core files (AGENTS, SOUL, USER, IDENTITY, TOOLS)60├── HEARTBEAT.md (if periodic tasks needed)61├── MEMORY.md (curated long-term memory)62├── docs/63│ ├── procedures/ (domain-specific workflows)64│ ├── standing-orders/ (autonomous programs)65│ └── [domain]/ (domain-specific reference docs)66└── memory/ (daily logs, auto-managed)67```6869## Quick Start: Minimal Viable Workspace7071For any use case, start with these three files:7273### 1. SOUL.md (under 500 words to start)74```markdown75## Core Truths76- [2-3 domain-specific values]77- Be resourceful before asking7879## Boundaries80- [2-3 hard limits specific to your domain]8182## Vibe83[One sentence describing communication style]84```8586### 2. AGENTS.md (under 1,000 words to start)87```markdown88## Session Startup891. Read SOUL.md, USER.md902. Read today's memory file9192## Safety93- Don't run destructive commands without confirmation94- Keep private data private9596## Memory97- Daily logs in memory/YYYY-MM-DD.md98- Long-term facts in MEMORY.md (main session only)99```100101### 3. USER.md (minimal profile)102```markdown103- Name: [your name]104- Timezone: [your timezone]105- Language: [preferred language]106- Style: [brief/detailed, formal/casual]107```108109Then iterate: add IDENTITY.md, TOOLS.md, HEARTBEAT.md, and docs/ as needs arise.
agents-store