claudeclaw

A lightweight, open-source OpenClaw version built into your Claude Code.

ClaudeClaw turns your Claude Code into a personal assistant that never sleeps. It runs as a background daemon, executing tasks on a schedule, responding to messages on Telegram and Discord, transcribing voice commands, and integrating with any service you need.

Note: Please don't use ClaudeClaw for hacking any bank system or doing any illegal activities. Thank you.

Why ClaudeClaw?

Category	ClaudeClaw	OpenClaw
Anthropic Will Come After You	No	Yes
API Overhead	Directly uses your Claude Code subscription	Nightmare
Setup & Installation	~5 minutes	Nightmare
Deployment	Install Claude Code on any device or VPS and run	Nightmare
Isolation Model	Folder-based and isolated as needed	Global by default (security nightmare)
Reliability	Simple reliable system for agents	Bugs nightmare
Feature Scope	Lightweight features you actually use	600k+ LOC nightmare
Security	Average Claude Code usage	Nightmare
Cost Efficiency	Efficient usage	Nightmare
Memory	Uses Claude internal memory system + CLAUDE.md	Nightmare

Getting Started in 5 Minutes

claude plugin marketplace add moazbuilds/claudeclaw
claude plugin install claudeclaw

Then open a Claude Code session and run:

/claudeclaw:start

The setup wizard walks you through model, heartbeat, Telegram, Discord, and security, then your daemon is live with a web dashboard.

What Would Be Built Next?

Mega Post: Help shape the next ClaudeClaw features. Vote, suggest ideas, and discuss priorities in this post.

Features

Automation

• Heartbeat: Periodic check-ins with configurable intervals, quiet hours, and editable prompts.
• Cron Jobs: Timezone-aware schedules for repeating or one-time tasks with reliable execution.

Communication

• Telegram: Text, image, and voice support.
• Discord: DMs, server mentions/replies, slash commands, voice messages, and image attachments.
• Time Awareness: Message time prefixes help the agent understand delays and daily patterns.

Reliability and Control

• GLM Fallback: Automatically continue with GLM models if your primary limit is reached.
• Web Dashboard: Manage jobs, monitor runs, and inspect logs in real time.
• Security Levels: Four access levels from read-only to full system access.
• Model Selection: Switch models based on your workload.

FAQ

Can ClaudeClaw do <something>?

If Claude Code can do it, ClaudeClaw can do it too. ClaudeClaw adds cron jobs, heartbeats, and Telegram/Discord bridges on top. You can also give your ClaudeClaw new skills and teach it custom workflows.

Is this project breaking Anthropic ToS?

No. ClaudeClaw is local usage inside the Claude Code ecosystem. It wraps Claude Code directly and does not require third-party OAuth outside that flow. If you build your own scripts to do the same thing, it would be the same.

Will Anthropic sue you for building ClaudeClaw?

I hope not.

Are you ready to change this project name?

If it bothers Anthropic, I might rename it to OpenClawd. Not sure yet.

Screenshots

Claude Code Folder-Based Status Bar