OpenLAIR/dr-claw

A Super AI Lab with massive AI Doctors as Assistants. Best IDE for Research via AI Power.

What it solves

Dr. Claw is a full-stack AI research workspace designed to streamline the end-to-end academic and technical research lifecycle. It eliminates the need to switch between disparate tools for literature surveys, ideation, experimentation, and paper writing, providing a unified environment where researchers can manage their entire pipeline from initial concept to final publication and promotion.

How it works

The system operates as a model-agnostic platform that integrates with various execution engines (such as Claude Code, Gemini CLI, Codex, and OpenRouter). It uses a "Research Lab" dashboard to structure projects into stages: Survey, Ideation, Experiment, Publication, and Promotion.

Key components include:

  • Agentic Execution: Agents use a library of over 100 built-in research skills to perform tasks like code surveys and statistical analysis.
  • Chat-Driven Pipelines: Users describe ideas in a chat interface, and the system generates a structured research brief and task list using a pipeline planner.
  • Auto Research Hub: Allows for one-click autonomous execution of research pipelines.
  • Integrated Tooling: Includes a built-in news dashboard for tracking arXiv and GitHub, a file/Git explorer, and a terminal for direct agent interaction.

Who it’s for

It is built for researchers and builders who want to accelerate their iteration speed and maintain scientific rigor while leveraging AI agents to handle the manual overhead of the research process.

Highlights

  • End-to-End Pipeline: Automates the flow from literature review (Survey) to manuscript drafting (Publication) and slide creation (Promotion).
  • Multi-Agent Backend: Supports a wide array of models via OpenRouter and native CLI tools from Anthropic, Google, and OpenAI.
  • Librarian of Skills: Features 100+ curated research skills that agents auto-discover and apply to tasks.
  • Unified Workspace: Combines a news feed, LaTeX-rendered ideas, Git source control, and a local GPU detection system in one interface.
  • Model Agnostic & Local: Runs on the user's own machine with their own data and GPUs, requiring no subscription.