ClawCore

AN AI ASSISTANT WITH A SOUL.

ClawCore extracts the soul of OpenClaw into a minimal, self-contained personal AI assistant. It keeps the personality system that makes AI feel alive, while stripping away the infrastructure complexity.

Get Started View on GitHub

›

Why ClawCore?

OpenClaw is powerful — but it's also complex. ClawCore asks: what if we keep only the soul?

›

Quick Start

Terminal

# Clone, install, run. That's it. 🦐

$ git clone https://github.com/dataelement/ClawCore.git
$ cd ClawCore
$ npm install
$ npm run dev

On first run, ClawCore will ask for your LLM API key, start a bootstrap conversation, and create your workspace.

›

Architecture

Clean, minimal, understandable. One file can change the world.

CLI (index.ts)

Agent (agent.ts)

📋

System Prompt Builder

Soul + Identity + Memory Index + Skills

🤖

LLM Provider

OpenAI-compatible API

🔧

Tool Executor

15 tools with permission enforcement

💓

Heartbeat Runner

setInterval with busy guard

›

Safe for Your Personal Machine

Most AI assistants with file access make people nervous. ClawCore solves this architecturally.

🛡️

File Safety

user/ is read-only. Processing happens in workbench/. The AI copies files before editing.

All file operations go through assertInsideWorkspace(), which resolves symlinks before checking paths.

⚡

Exec Safety

Shell commands use three layers: whitelist (safe), blocklist (dangerous), and confirmation (unknown).

✅ Allowed: ls, cat, grep, wc, open

🚫 Blocked: rm, curl, wget, sudo, ssh

⚠️ Confirm: python3, node, etc.