openclaw-team

Clone it. Install OpenClaw. Add an agent. Ask it to turn this on.

A production-ready template for running a 5-agent OpenClaw team with shared project management, inter-agent communication via Convex, local search via SearXNG, and automated cron-based agile workflows.

🚀 Quick Start

Prerequisites

• OpenClaw installed (npm install -g openclaw@latest)
• Docker installed and running

(see docker/README.md for install instructions)

• An LLM API key (Anthropic recommended)
• Node.js 22+

Deploy Your Own

Edit .env and add your ANTHROPIC_API_KEY (and optionally OPENAI_API_KEY, etc.).

• Make it yours — edit openclaw.json5 directly. See Customization below for names, models, etc.

See docker/README.md and convex/README.md for detailed setup, troubleshooting, and running as persistent services.

That's it. Just start a chat and tell your Product Lead agent: "Read the README and get the team running."

💡 Note: You can run openclaw onboard if you like — it won't overwrite openclaw.json5. Everything it configures (API keys, models, daemon) is already handled here via .env and openclaw.json5.

🔌 Services

This template depends on two external services. Both are designed to run persistently and survive reboots.

SearXNG (local search)

A private, self-hosted metasearch engine. Runs as a Docker container.

Convex (database)

Cloud-hosted real-time database for task management and inter-agent messaging.