AcaClawAcaClawHub
The academic distribution of OpenClaw

Your own AI
co-scientist.

Whether you need an AI biologist, AI chemist, or data scientist — AcaClaw gives you discipline-specific digital co-scientists, ready out of the box. Like Ubuntu is to Linux, AcaClaw is to OpenClaw: everything you need, one install.

Many
AI agents
40+
Curated skills
3-tier
Security
100%
Tested & reviewed
AcaClaw project poster

What AcaClaw adds to OpenClaw

Built on the open-source OpenClaw platform, AcaClaw adds everything researchers need — no terminal required.

User-friendly GUI

No terminal knowledge needed. Manage skills, chat with agents, browse backups, and configure everything through an intuitive browser interface.

Workspace & project system

All file modifications and deletions happen inside the workspace by default — preventing accidental data loss and keeping your system clean.

Academic skills

Discipline-specific skills deeply tailored to your field — paper search, data analysis, citation management, figure generation, and more.

Computing environment

Discipline-specific Conda environments with unified management of Python, R, CUDA, and scientific computing dependencies. Skills are guaranteed compatible.

Data backup

Every modified file is automatically backed up with versioned snapshots and SHA-256 integrity checks. Your data is always safe and recoverable.

One install, everything ready

Like Ubuntu is to Linux — all skills pre-installed, all environments pre-configured, all agents ready. Install, pick your discipline, start working.

A team of AI co-scientists

Each agent has its own persona, expertise, Conda environment, and skill set. Run them in parallel — ask multiple agents simultaneously.

🧬

Dr. Gene

Biology

Genomics, RNA-seq, phylogenetics, protein structure

🏥

Dr. Curie

Medicine

Clinical trials, survival analysis, epidemiology

🤖

Dr. Turing

AI / ML

Neural networks, NLP, deep learning, benchmarking

📊

Dr. Bayes

Statistics

Pandas, R/tidyverse, visualization, hypothesis testing

💻

Dr. Knuth

Computer Science

Algorithms, systems, code review, architecture

Isolated workspace per agent
Independent parallel chat tabs
Shared data directory for collaboration

Your data stays safe

Three-tier security model with automatic backup. Every file is versioned before modification — nothing is ever lost.

Default

Workspace confinement

All AI operations restricted to ~/AcaClaw/. Tool deny-lists, command deny-lists, network allowlists, credential scrubbing, prompt injection detection.

Opt-in

Remote access

TLS required, auth token mandatory. Supports loopback, Tailscale, and LAN bind modes. Security alert before enabling.

Maximum

Docker sandbox

All code runs inside a container with dropped capabilities, read-only root filesystem, no network, seccomp profile, non-root execution.

Auto-backup before every modification
SHA-256 integrity verification
Full audit trail of every action
Privacy-first — no telemetry

One command to install

Installs OpenClaw, AcaClaw plugins, Miniforge, and all academic skills. Then opens a browser wizard to pick your discipline and connect your AI provider.

macOS (Intel & Apple Silicon)
$ curl -fsSL https://acaclaw.com/install.sh | bash
Linux (x86_64 & arm64)
$ curl -fsSL https://acaclaw.com/install.sh | bash

The installer handles everything automatically — Node.js, Conda, and all dependencies.

01

Install

One command installs everything

02

Pick discipline

Biology, Chemistry, Medicine, Physics, or General

03

Connect AI

OpenAI, Google, Anthropic, or OpenClaw Web

04

Start working

Browser opens — no terminal needed

The distribution

A pre-shipped build of OpenClaw with every curated skill installed and tested in discipline-specific Conda environments. Install, pick your field, and start working.

  • All skills pre-installed — zero setup
  • Discipline Conda envs — no dependency conflicts
  • macOS, Linux, and Windows (x86_64 + arm64)
  • Security-by-default with workspace confinement

The skills platform

Browse, search and contribute AI skills across every academic discipline. Every skill is team-built, tested, and individually credited — the academic model for open-source AI.

  • 10 disciplines + cross-discipline academic skills
  • 7 recognized roles — creator to documenter
  • Security review, env compatibility, CI tests
  • Published to ClawHub under @acaclaw

Browse by discipline

Skills organized by academic field — find exactly what your research needs.

From idea to your machine

Every skill goes through a rigorous pipeline before it ships in AcaClaw. The academic model — the journal publishes, the authors get credit.

01

Develop

Contributors build skills with specific roles — creator, author, tester, maintainer.

02

Test

CI tests, security scan, env compatibility check. Must pass all quality gates.

03

Publish

Published to ClawHub under @acaclaw. Every contributor credited by name and role.

04

Ship

Bundled into AcaClaw. Users install once and every skill is pre-installed, pre-tested, ready.

7 recognized contributor roles

Creator
Tester
Reviewer
Maintainer
+3 more

Built for every researcher

No coding required. Describe what you need in plain language.

🎓

Graduate students

Search papers, manage references, draft manuscripts

🔬

Lab researchers

Process data, generate figures, write reports

👩‍🏫

Professors & PIs

Draft grants, review manuscripts, prepare courses

📚

Students

Literature review, lab reports, step-by-step help

Get AcaClaw

Install the complete academic distribution. All skills pre-installed, one environment, ready in minutes.

Install

Start contributing

Create skills, test them, fix bugs, review code, or write docs. Every contribution is tracked. Every contributor is named.