DEV Community

WonderLab
WonderLab

Posted on

Open Source Project of the Day (Part 24): OpenClawInstaller - One-Click Deployment for Your Private AI Assistant OpenClaw

#openclaw #opensource #ai

Introduction

"Put Claude/GPT into Telegram, Discord, and Feishu — one command handles environment detection, installation, configuration, and startup."

This is Part 24 of the "Open Source Project of the Day" series. Today we explore OpenClawInstaller (GitHub), maintained by miaoxworld.

OpenClaw (official site clawd.bot) is a self-hostable AI assistant gateway: it connects to Claude, GPT, Gemini, Ollama, and other models, lets you chat through Telegram, Discord, WhatsApp, Feishu, Slack, WeChat, and other channels, and offers persistent memory, proactive notifications, a skill system, and optional remote control. But setting up OpenClaw from scratch — configuring the Node environment, choosing models, setting up channels — involves many steps. OpenClawInstaller was built for exactly this: a one-click install script + interactive configuration menu that automatically handles dependency detection, OpenClaw installation, AI model and messaging channel configuration, API testing, and service startup — and can be paired with the OpenClaw Manager desktop app for visual management.

What You'll Learn

  • OpenClawInstaller's purpose: lowering the barrier to deploying and configuring OpenClaw
  • Supported AI models (Claude/GPT/Gemini/OpenRouter/Groq/Mistral/Ollama) and messaging channels (Telegram/Discord/Feishu/WhatsApp, etc.)
  • Quick start: one-click curl install vs. manual install, post-install startup and config menu
  • Common commands (service management, configuration, backup) and security recommendations
  • Relationship to the main OpenClaw repository and OpenClaw Manager desktop app

Prerequisites

  • Basic command-line operations (bash, environment variables)
  • For cloud models: you'll need the corresponding API key (Anthropic/OpenAI, etc.); for Ollama, it must be installed locally
  • For Telegram/Discord/Feishu, etc.: you'll need to create an app or bot on the respective platform and obtain the Token/ID

Project Background

Project Introduction

OpenClawInstaller is a one-click deployment tool for OpenClaw (ClawdBot), providing:

  1. Install script (install.sh): Detects the system (macOS / Ubuntu / Debian / CentOS), installs dependencies and Node environment, installs OpenClaw (openclaw) via npm, guides AI model and identity configuration, tests the API, optionally auto-starts the Gateway and opens the configuration menu
  2. Configuration menu (config-menu.sh): Interactive configuration for AI models, messaging channels (Telegram, Discord, WhatsApp, Feishu, etc.), advanced settings, and updates
  3. Documentation and best practices: README includes system requirements, quick start, multi-model/channel configuration instructions, security recommendations, and FAQs

After deployment, the OpenClaw Gateway runs as a background service, and users can chat in Telegram, Discord, and other channels with the same AI capabilities, enjoying persistent memory, the skill system, and more.

Core problems the project solves:

  • OpenClaw is feature-rich but the initial deployment involves many steps, easily getting stuck on environment or configuration
  • Need a single place summarizing "multi-model, multi-channel" configuration guidance and executable scripts
  • Want an "install-and-use" experience, while retaining the ability to fine-tune later via menu or desktop app

Target user groups:

  • Users who want to quickly get OpenClaw running on their own machine (individuals or small teams)
  • Developers or operators needing to integrate Telegram, Discord, Feishu, WhatsApp, etc.
  • Users who prefer a graphical interface (can pair with OpenClaw Manager desktop app)

Author/Team Introduction

  • Maintainer: miaoxworld (GitHub), also maintains OpenClaw Manager desktop management tool (Tauri 2.0 + React + TypeScript + Rust)
  • Upstream project: OpenClaw main repository openclaw/openclaw
  • Project creation date: January 2026 (README changelog shows v1.0.0 2026-01-29; GitHub shows repository created 2026-01-29)

Project Stats

  • ⭐ GitHub Stars: 1.7k+
  • 🍴 Forks: 265+
  • 📦 Version: v1.0.0 (first release 2026-01-29)
  • 📄 License: MIT (noted in README)
  • 🌐 OpenClaw website/docs: clawd.bot, clawd.bot/docs
  • 💬 Community: GitHub Issues, Discussions (if available)

Main Features

Core Purpose

OpenClawInstaller's core purpose is to make OpenClaw installation and initial configuration repeatable and low-effort:

  1. Environment and dependencies: Detects OS and Node version (requires Node.js v22+), guides or automatically installs dependencies
  2. Install OpenClaw: Installs the global CLI and Gateway via npm install -g openclaw
  3. Guided configuration: Interactively select AI model (including custom API endpoints), enter API keys, optionally configure messaging channels
  4. Validate and start: Test API connection, optionally start Gateway, provide configuration menu entry for later fine-tuning

Use Cases

  1. Personal private AI assistant

    • One-click deploy on a VPS or local machine, chat with Claude or GPT via Telegram/Discord, with full control over data and keys

  2. Internal team bot

    • Integrate with Feishu/Slack, share the same model and memory across the team, use the skill system for internal tool encapsulation

  3. Multi-model and API relay

    • Supports custom API endpoints like OneAPI/NewAPI for relay or self-built API; also supports OpenRouter, Groq, Ollama, etc.

  4. Quick OpenClaw experience

    • If you're unfamiliar with Node or OpenClaw configuration, use one curl command to go from zero to working

  5. Paired with desktop management

    • After installation, use OpenClaw Manager for visual monitoring, configuration, and updates

Quick Start

System requirements: macOS 12+ / Ubuntu 20.04+ / Debian 11+ / CentOS 8+; Node.js v22+; 4GB+ RAM recommended; at least 1GB disk space.

Method 1: One-click install (recommended)

curl -fsSL https://raw.githubusercontent.com/miaoxworld/OpenClawInstaller/main/install.sh | bash
Enter fullscreen mode Exit fullscreen mode

The script detects your environment, installs OpenClaw, guides model and identity configuration, tests the API, and asks whether to start the service. You can optionally run the configuration menu afterward to configure channels.

Method 2: Manual install

git clone https://github.com/miaoxworld/OpenClawInstaller.git
cd OpenClawInstaller
chmod +x install.sh config-menu.sh
./install.sh
Enter fullscreen mode Exit fullscreen mode

If macOS permissions cause installation failure, manually run npm install -g openclaw first, then run ./install.sh.

Common commands after installation

# Start service (background)
openclaw gateway start

# Open configuration menu (channels, models, advanced settings, etc.)
bash ~/.openclaw/config-menu.sh

# Or run the latest config menu via curl
curl -fsSL https://raw.githubusercontent.com/miaoxworld/OpenClawInstaller/main/config-menu.sh | bash
Enter fullscreen mode Exit fullscreen mode

Desktop app (GUI): If you prefer a graphical interface, use OpenClaw Manager (Tauri + React) — supports real-time monitoring, visual model and channel configuration, macOS/Windows/Linux.

Core Features

  1. Multi-model support

    • Anthropic Claude (supports custom API endpoint), OpenAI GPT (supports custom API, requires Responses API support), Google Gemini, OpenRouter, Groq, Mistral AI, Ollama (local, no API key needed)

  2. Multi-channel integration

    • Telegram, Discord, WhatsApp, Slack, WeChat, iMessage (macOS only), Feishu; configuration menu provides step-by-step instructions per channel (Bot Token, Channel ID, Feishu App ID/Secret, etc.)

  3. Interactive configuration menu

    • AI model configuration, messaging channel configuration, API testing, diagnostics, and updates all in one menu — no need to manually edit configuration files

  4. OpenClaw core capabilities

    • Persistent memory, proactive notifications (scheduled reminders, morning briefings, etc.), skill system (capabilities defined in Markdown), optional remote control (enable with caution and read security guidelines)

  5. Cross-platform

    • Supports macOS, Ubuntu, Debian, CentOS; desktop app supports macOS, Windows, Linux

Project Advantages

Comparison OpenClawInstaller + OpenClaw Manual OpenClaw deployment Channel-specific bots + API only
Deployment speed One command + guided config Must read docs and configure environment manually No unified gateway or memory
Multi-model/channel Unified gateway configuration Same but manual Need separate integration per channel
Persistent memory & skills Built into OpenClaw Same Usually absent or requires custom build
GUI management Optional OpenClaw Manager Same Product-dependent

Why choose OpenClawInstaller?

  • Lower barrier: From "don't know how to install OpenClaw" to "one command + follow prompts to fill in keys"
  • Pairs with main repo: Installs the official openclaw; version and capabilities match openclaw/openclaw — the Installer only handles the install and config flow
  • Optional desktop app: CLI and OpenClaw Manager can coexist — choose based on preference

Detailed Project Analysis

Installation Process Overview

  1. Run install script: install.sh detects OS, checks/installs Node (v22+), installs openclaw if needed (npm install -g openclaw)
  2. Environment and configuration: Config is written to ~/.openclaw/env (API Key, BASE_URL, etc.) and ~/.openclaw/openclaw.json (model, channels, etc.); custom API endpoints are written as custom providers in openclaw.json
  3. Guidance and testing: Interactively select model, enter key, optionally test API; ask whether to start Gateway, whether to open config menu
  4. Ongoing management: Manage via openclaw gateway start/stop/restart, openclaw config, openclaw doctor, config-menu.sh, or OpenClaw Manager

Configuration and Directory Structure

  • Environment variables: ~/.openclaw/env, e.g., ANTHROPIC_API_KEY, ANTHROPIC_BASE_URL (or OpenAI, etc.)
  • OpenClaw config: ~/.openclaw/openclaw.json (managed by OpenClaw and the install/config flow)
  • Other: ~/.openclaw/backups/, ~/.openclaw/logs/, etc.

It's recommended to use the install wizard or config-menu.sh rather than directly editing the JSON to avoid errors.

Security Recommendations (README Highlights)

  • Permissions: OpenClaw can be configured to execute system commands, access files, etc.; by default it's recommended to disable this or only enable it in controlled environments; avoid running with elevated permissions long-term on your main work machine — use a dedicated machine or VM
  • Sensitive information: API keys, tokens, etc. should be stored in environment variables and never committed to public repositories
  • Channel permissions: Restrict Telegram and similar channels to respond only to specific User IDs to reduce misuse and abuse

Common Issues Quick Reference

  • Node version too low: Upgrade to v22+ (e.g., macOS: brew install node@22, Ubuntu/Debian: NodeSource setup_22.x)
  • Can't connect after startup: Run openclaw doctor, check openclaw logs
  • Telegram not responding: Verify Bot Token, User ID, and allowed_users (if enabled), network/proxy
  • Update: npm update -g openclaw or use "Update OpenClaw" in the configuration menu
  • Uninstall: openclaw gateway stop → npm uninstall -g openclaw → optionally rm -rf ~/.openclaw

Project Resources

Official Resources

Related Resources

  • Repo README: System requirements, quick start, multi-model/channel configuration, security recommendations, FAQs, changelog
  • Feishu configuration: README points to docs/feishu-setup.md (if available) for Feishu bot configuration and persistent connection instructions
  • Plugins like Claude-Mem can be used with the OpenClaw gateway (see Claude-Mem documentation for OpenClaw integration)

Who Should Use This

  • Users who want to get OpenClaw running quickly: One command for installation and basic configuration
  • Developers or teams needing Telegram/Discord/Feishu/WhatsApp channel integration: Script and documentation covers each channel's steps
  • Those who prefer a GUI: Use OpenClaw Manager for daily management and monitoring after installation
  • Self-built AI gateways and multi-model setups: Supports official API and custom API endpoints, easy to combine with existing proxies or self-built services

Welcome to visit my personal homepage for more useful knowledge and interesting products

Top comments (0)