ScienceClaw is a personal research assistant built with LangChain DeepAgents and AIO Sandbox infrastructure, adopting a completely new architecture beyond OpenClaw. It offers stronger security, better transparency, and a more user-friendly experience.
# Add to your Claude Code skills
git clone https://github.com/AgentTeam-TaichuAI/ScienceClawLast scanned: 5/14/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-05-14T06:46:23.952Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}ScienceClaw is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by AgentTeam-TaichuAI. ScienceClaw is a personal research assistant built with LangChain DeepAgents and AIO Sandbox infrastructure, adopting a completely new architecture beyond OpenClaw. It offers stronger security, better transparency, and a more user-friendly experience. It has 544 GitHub stars.
Yes. ScienceClaw passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/AgentTeam-TaichuAI/ScienceClaw" and add it to your Claude Code skills directory (see the Installation section above).
ScienceClaw is primarily written in Python. It is open-source under AgentTeam-TaichuAI on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh ScienceClaw against similar tools.
No comments yet. Be the first to share your thoughts!
ScienceClaw is a personal research assistant built with LangChain DeepAgents and AIO Sandbox infrastructure, adopting a completely new architecture beyond OpenClaw. It offers stronger security, better transparency, and a more user-friendly experience.
1,900+ built-in scientific tools · Multi-format content generation · Fully local & privacy-first
Why ScienceClaw · Architecture · News · Quick Start · Demo · Free API Credits · Tools & Skills · Features · Project Structure · Commands · Community · Acknowledgements
ScienceClaw runs entirely inside Docker containers. The agent cannot access your host system, personal files, or environment variables. All code execution happens in an isolated sandbox, and generated data stays in a local ./workspace directory — nothing is uploaded to external servers. Deploy with confidence on your own machine.
Every step of the agent's workflow is visible and traceable — from web search and data crawling, to reasoning and tool invocation, to final report generation. You always know where results come from, what actions were taken, and how conclusions were reached. Inspect any step at any time.
No tedious configuration needed. ScienceClaw ships with curated tools and skill packages — launch the entire environment with a single command. Whether you're a researcher, developer, or student, you can get started immediately. Focus on real tasks, not setup.
Welcome to try our official cloud version: scienceclaw.zidongtaichu.com
.edu or .edu.cn email and get an extra ¥50 in trial creditsNo Docker required, no command-line needed. Download the desktop installer and get started instantly.
1. Download the installer
👉 ScienceClaw Desktop v0.0.4 (.tar.gz)
2. Extract and install
After downloading, extract the archive and run the installer. Follow the setup wizard to complete installation.
3. Launch
Once installed, double-click the desktop shortcut to start ScienceClaw — ready to use out of the box.
📖 For a detailed step-by-step deployment guide, see Deployment Guide (中文).
1. Get the code
git clone https://github.com/AgentTeam-TaichuAI/ScienceClaw.git
cd ScienceClaw
cd ScienceClaw
git pull
2. Launch — pull pre-built images
docker compose -f docker-compose-release.yml up -d --pull always
Pulls pre-built images directly — no local compilation needed. Ready in a few minutes.
3. Open in browser
http://localhost:5173
4. Login
docker compose up -d --build
Builds all images from source code. Ideal for developers who need to modify the code. The first build downloads dependencies and may take longer.
To lower the barrier for new users, a limited batch of LLM API resources:
| Offer | Details |
|---|---|
| SCNet (National Supercomputing Internet) | 10M free tokens (Claim here) |
| Zidong Taichu Cloud | 10M free tokens (Claim here) |
Limited availability — first come, first served. We will continue to secure more compute resources for the community.
ScienceClaw integrates ToolUniverse, a unified ecosystem of 1,900+ scientific tools spanning multiple disciplines:
| Domain | Capabilities |
|---|---|
| 💊 Drug Discovery & Biomedicine | Target identification (OpenTargets), ADMET prediction, drug safety (FAERS), protein analysis (UniProt, PDB, AlphaFold), genomics (GWAS, GTEx), clinical trials |
| 🔭 Astronomy & Space Science | SIMBAD astronomical objects, SDSS sky survey, NASA exoplanet archive, JPL Horizons ephemeris, NASA DONKI solar events, small body database |
| 🌍 Earth & Environmental Science | USGS earthquakes & hydrology, ERDDAP ocean/climate data, SoilGrids, air quality (WAQI), OpenMeteo weather/climate, marine regions |
| ⚗️ Chemistry & Materials | COD crystal structures, molecular property prediction, SMILES-based analysis, compound similarity, chemical computation |
| 🌱 Biodiversity & Ecology | GBIF species records, OBIS marine biodiversity, POWO plant taxonomy, WoRMS marine species, eBird taxonomy, paleobiology database |
| 📊 Social Science & Statistics | World Bank indicators, Eurostat, US Census population data, Wikidata knowledge graph, DBpedia |
| 📚 Academic Literature | Multi-source search (PubMed, arXiv, OpenAlex, Semantic Scholar, DBLP, INSPIRE-HEP, Crossref, DOAJ, CORE) |
| 🤖 Data Science & Computing | HuggingFace models/datasets, OpenML, GitHub repositories, scientific computing software, image processing |
| Layer | Description | Examples |
|---|---|---|
| 🔧 Built-in Tools | Core search & crawl capabilities | web_search, web_crawl |
| 🧪 ToolUniverse | 1,900+ scientific tools, ready to use | UniProt, OpenTargets, FAERS, PDB, ADMET, etc. |
| 📦 Sandbox Tools | File operations & code execution | read_file, write_file, execute, shell |
| 🛠️ Custom @tool | User-defined Python functions, hot-loaded from Tools/ |
Your own tools |
ScienceClaw makes it easy to extend with your own tools:
@tool decorated functions into the Tools/ directory; they are auto-detected and hot-loaded without restart.Skills are structured instruction documents (SKILL.md) that guide the Agent through complex, multi-step workflows. Unlike tools (executable code), skills act as the Agent's "playbook" — defining strategies, rules, and best practices.
| Skill | Purpose |
|---|---|
| Read, create, merge, split, OCR, and generate professional PDF research reports | |
| 📝 docx | Create and edit Word documents with cover pages, TOC, tables, and charts |
| 📊 pptx | Generate and edit PowerPoint slide decks |
| 📈 xlsx | Create and manipulate Excel spreadsheets, CSV/TSV data processing |
| 🛠️ tool-creator | Create & upgrade custom @tool functions (write → test → save) |
| 📝 skill-creator | Create & refine skills with draft → test → review → iterate |
| 🔍 find-skills | Search, discover & install community skills from the open ecosystem |
| 🧪 tooluniverse | Unified access to 1,900+ scientific tools |
ScienceClaw can produ