Dr. Claw is a full-stack AI research workspace for researchers and builders, accelerating end-to-end projects from idea to publication. Addressing the slowness of manual work and the generality of fully automated AI, it enables "Leveraged Cognition" and "Agentic Execution," boosting iteration speed and research quality while maintaining scientific rigor.

How It Works

Uses "Leveraged Cognition" and "Agentic Execution" to transform research ideas into actionable pipelines. It structures research through a Survey → Ideation → Experiment → Publication → Promotion workflow. Tasks are generated interactively via chat, creating structured briefs and task lists. This approach balances human insight with AI capabilities for efficient, rigorous research.

Quick Start & Requirements

Clone (git clone https://github.com/OpenLAIR/dr-claw.git), cd dr-claw, npm install. Configure .env.example to .env. Run with npm run dev. Requires Node.js v20+ (v22 LTS recommended) and Claude Code, Gemini CLI, or Codex CLI. Native build tools may be needed.

Highlighted Details

• Research Lab: Dashboard for end-to-end research tracking and artifact inspection.
• Auto Research: One-click sequential task execution with email notifications.
• 100+ Research Skills: Library for diverse research tasks, auto-discovered by agents.
• Chat-Driven Pipeline: Interactive generation of research briefs and task lists via chat.
• Multi-Agent Backend: Supports Claude Code, Gemini CLI, and Codex.
• Interactive Chat + Shell: Side-by-side chat and terminal access.
• File & Git Explorer: In-app file management and Git operations.
• Responsive & PWA-Ready: Optimized for desktop, tablet, and mobile.

Maintenance & Community

Community engagement is encouraged via starring and watching the repository. Contributions are welcomed via pull requests. Specific community channels are not detailed.

Licensing & Compatibility

Dual-licensed: GPL-3.0 (upstream Claude Code UI) and AGPL-3.0 (original Dr. Claw contributions). AGPL-3.0 imposes strong copyleft obligations, especially for network services. Commercial use requires careful review.

Limitations & Caveats

Cursor agent support is in progress. AGPL-3.0 has implications for network deployments. Agent permissions require careful configuration to mitigate security risks, as permissive modes can grant broad access.