Complete, production-ready template for spinning up an OpenClaw instance with preconfigured workspace, agents, and automations.
Use case: Deploy on any VPS or local machine and have a fully operational OpenClaw environment in <30 minutes.
- Linux/macOS (WSL2 on Windows supported)
- Node.js 20+ (for OpenClaw CLI)
- Git
- 2GB disk space minimum
# 1. Install OpenClaw (if not already installed)
npm install -g openclaw
# 2. Clone this repo
git clone https://github.com/ainmissioncontrol/open-claw.git
cd open-claw
# 3. Run setup
bash scripts/setup.sh
# 4. Configure credentials (see CREDENTIALS_SETUP.md)
bash scripts/credentials-setup.sh
# 5. Start
openclaw gateway restartDone. Your OpenClaw instance is now running with:
- ✅ All workspace files loaded
- ✅ Cron jobs configured
- ✅ Agents ready
- ✅ Skills installed
- ✅ Memory system active
- SOUL.md — Jessica (COO) role definition
- AGENTS.md — Safety rules, team structure
- CAPABILITIES.md — What Jessica can/can't do
- MEMORY.md — Three-tier memory system (searchable recall)
- HEARTBEAT.md — Automated health checks every 30 min
- TOOLS.md — Complete toolkit reference
- research-bot — Content research & keyword clustering
- copywriter — Draft LinkedIn content in Jeff's voice
- editor — Final review & compliance check
- content-manager — Publishing & distribution
- builder — Infrastructure & deployment
- atlas — Prospecting & outreach
- Heartbeat checks (30-min intervals)
- Nightly checkpoints (memory compaction)
- Morning briefs
- Email digests
clawhub— Skill discovery & managementgog— Google Workspace integrationweather— wttr.in weather APIvideo-frames— ffmpeg video extractionhimalaya— Email management (IMAP/SMTP)tmux— Remote terminal control- Custom skills in
workspace/skills/
See CREDENTIALS_SETUP.md for step-by-step guide:
- GitHub PAT
- Mission Control API token
- Neon PostgreSQL
- Telegram API keys
- Google Workspace credentials
- Optional: GoHighLevel, ClickUp, Supabase
Edit workspace/SOUL.md to change:
- Agent name (replace "Jessica" with your name)
- Role title & responsibilities
- Communication rules
- Model preferences
Copy your own files into:
workspace/reference/— Knowledge baseworkspace/agents/*/SOUL.md— Agent customizationsworkspace/memory/— Daily logs & archived context
open_claw/
├── README.md (this file)
├── .gitignore
├── CREDENTIALS_SETUP.md
├── install.sh (automated setup)
│
├── scripts/
│ ├── setup.sh — Post-install configuration
│ ├── bootstrap.sh — Workspace initialization
│ ├── credentials-setup.sh — Interactive credential entry
│ └── health-check.sh — Verify system status
│
├── config/
│ ├── openclaw.json — Gateway configuration template
│ ├── cron-jobs.json — Scheduled tasks
│ └── agents/ — Agent SOUL.md templates
│
├── workspace/
│ ├── SOUL.md — COO role definition
│ ├── AGENTS.md — Safety & team rules
│ ├── CAPABILITIES.md — Capability matrix
│ ├── MEMORY.md — Memory system starter
│ ├── TOOLS.md — Toolkit reference
│ ├── HEARTBEAT.md — Health check config
│ │
│ ├── reference/
│ │ ├── VOICE.md — Writing voice & style
│ │ ├── CONTENT.md — Content strategy
│ │ ├── COMPLIANCE.md — Legal/regulatory
│ │ ├── BUSINESS.md — Business model
│ │ ├── BRAIN.md — Knowledge base
│ │ └── ...more reference docs
│ │
│ ├── agents/
│ │ ├── research-bot/SOUL.md
│ │ ├── copywriter/SOUL.md
│ │ ├── editor/SOUL.md
│ │ ├── content-manager/SOUL.md
│ │ ├── builder/SOUL.md
│ │ └── atlas/SOUL.md
│ │
│ ├── skills/ — Custom agent skills
│ │
│ ├── memory/
│ │ └── README.md — Memory architecture
│ │
│ └── content-pipeline/
│ ├── research/
│ ├── outlines/
│ ├── drafts/
│ ├── approved/
│ └── published/
│
└── docker/ (optional)
├── Dockerfile
└── docker-compose.yml
openclaw gateway restart# Spawn a sub-agent
openclaw agent spawn research-bot --session isolated --model claude-sonnet
# Send message to agent
echo "Research topic X" | openclaw agent message research-botopenclaw status
bash scripts/health-check.shtail -f ~/.openclaw/agents/main/sessions/*.jsonl- Edit
workspace/SOUL.md— replace "Jessica" with your name - Edit
workspace/AGENTS.md— update role description - Edit agent SOUL files in
workspace/agents/*/
Place .md files in workspace/skills/ following this structure:
workspace/skills/my-skill/
├── SKILL.md (main documentation)
├── examples/ (usage examples)
└── references/ (supporting docs)
Edit config/cron-jobs.json to:
- Change heartbeat interval
- Add new scheduled tasks
- Adjust notification channels
# Check logs
openclaw status
# Try restart
openclaw gateway restart
# Manual health check
bash scripts/health-check.sh# Verify skills are installed
openclaw skills list
# Check agent sessions
openclaw sessions list
# Restart gateway
openclaw gateway restart# Verify memory files exist
ls -la ~/.openclaw/workspace/memory/
ls -la ~/.openclaw/workspace/MEMORY.md
# Check file permissions
chmod 644 ~/.openclaw/workspace/memory/*- Never commit
CREDENTIALS.md(in.gitignore) - Store credentials in
.envlocally (not in repo) - All tokens should be rotated every 90 days
- Use GitHub's credential scanning:
gh secret scan - Review agent access logs:
~/.openclaw/agents/main/sessions/
- Docs: https://docs.openclaw.ai
- Discord: https://discord.com/invite/clawd
- GitHub: https://github.com/openclaw/openclaw
- Skill Hub: https://clawhub.com
These templates and customizations are provided as-is. OpenClaw core is MIT licensed.
Last updated: 2026-03-11
Status: Production-ready
Contributors: Angel Investors Network team