# ScienceClawPlus
**Repository Path**: elfbobo_admin_admin/science-claw-plus
## Basic Information
- **Project Name**: ScienceClawPlus
- **Description**: ScienceClaw 是一个科学研究智能体。它搜索文献、查询数据库、运行分析、生成图表、撰写报告——零自定义代码，完全基于 OpenClaw 构建，仅靠一个 Markdown 文件（SCIENCE.md，约 200 行）和 264 个领域技能。
本项目原始版本来源于：GitHub的Zaoqu-Liu/ScienceClaw，后续改进版本由本项目完成。
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No
## Statistics
- **Stars**: 0
- **Forks**: 1
- **Created**: 2026-03-17
- **Last Updated**: 2026-03-26
## Categories & Tags
**Categories**: Uncategorized
**Tags**: None
## README
# ScienceClaw
**Your AI Research Colleague**
[](LICENSE)
[](https://gitee.com/XiaoYRecluse/science-claw-plus/releases)
[](#-skills)
[](#-database-access)
[](#-deep-research)
[](https://nodejs.org/)
**EN** | [中文](docs/i18n/README_ZH.md) | [日本語](docs/i18n/README_JA.md) | [한국어](docs/i18n/README_KO.md)
---
> **🆕 ScienceClawPlus v1.1.0** — This is an enhanced version based on [ScienceClaw](https://github.com/Zaoqu-Liu/ScienceClaw) with additional features:
> - ✨ Custom model support (DeepSeek, Qwen, ERNIE, Ollama)
> - ✨ Feishu/Lark integration
> - ✨ Interactive configuration commands
> - ✨ Optimized for China users
>
> See [CHANGELOG.md](CHANGELOG.md) for details.
---
ScienceClaw is a science research agent. It searches literature, queries databases, runs analyses, generates figures, and writes reports. Zero custom code — built entirely on [OpenClaw](https://github.com/openclaw/openclaw) with one markdown file (`SCIENCE.md`, ~200 lines) and 264 domain skills. The model does 99% of the work; the markdown teaches it how to be a scientist.
---
## See It In Action
### Case 1 — Investigate the role and significance of THBS2 in tumors
> **Prompt:** *"Investigate the role and significance of THBS2 in tumors"*
ScienceClaw autonomously searched PubMed, queried TCGA via cBioPortal and TIMER2.0, ran survival analyses in R, and compiled a 30-page report with 87 citations.
**Key findings:**
- THBS2 is significantly upregulated in **17 out of 33 TCGA cancer types**
- Combined THBS2 + CA19-9 panel achieved diagnostic AUC of **0.96** in a retrospective pancreatic cancer cohort — but dropped to **0.69** in a prospective validation set
- Tumor microenvironment analysis revealed THBS2 correlation with M2 macrophage infiltration across multiple cancer types
[Read the full case study →](docs/cases/case-thbs2-tumor.md)
---
### Case 2 — Survey the applications of LLM in biomedicine
> **Prompt:** *"Survey the applications of LLM in biomedicine"*
ScienceClaw conducted a systematic literature search across PubMed, Semantic Scholar, and OpenAlex, then synthesized findings into a structured survey with trend analysis and visualizations.
**Key findings:**
- Medical LLM publications grew **570x in two years** — from 8 in 2022 to 4,562 in 2024
- Med-PaLM 2 reached **86.5%** accuracy on USMLE, surpassing the expert physician threshold
- The healthcare LLM market is projected to reach **$110B by 2030**
[Read the full case study →](docs/cases/case-llm-biomedicine.md)
---
## Quick Start
### Prerequisites
| Requirement | Version | Notes |
|-------------|---------|-------|
| Node.js | >= 22 | Required |
| pnpm | Latest | Required — `npm install -g pnpm` |
| Python | >= 3.10 | For code execution (R, Julia optional) |
| Docker | Latest | Optional — for sandboxed code execution |
### Step 1 — Clone and setup
```bash
git clone https://gitee.com/XiaoYRecluse/science-claw-plus.git
cd science-claw-plus # Navigate to the project directory
chmod +x scienceclaw # Add executable permission to the script
bash scripts/setup.sh # auto-starts gateway + opens TUI
```
> **Note:** The `./scienceclaw` script must be run from the project directory (`science-claw-plus/`). It will not work if run from a different location.
> **China users:** When setup asks for API key, use [DeepSeek](https://platform.deepseek.com/) (no proxy needed, very affordable) or [yunwu.ai](https://yunwu.ai/) relay. See the [Installation Guide](docs/getting-started/installation.md) for details.
### Step 2 — Run
```bash
./scienceclaw run # auto-starts gateway + opens TUI
```
That's it. Two commands.
### Having model issues?
```bash
./scienceclaw models # check which models work, diagnose 404/403 errors
./scienceclaw doctor # full system health check
```
### Add or switch models (optional)
ScienceClaw supports a wide range of LLM providers. You can easily add domestic models, local deployments, or custom endpoints:
```bash
./scienceclaw add model # interactive wizard to add new models
```
**Supported model providers:**
| Category | Provider | Models |
|----------|----------|--------|
| **International** | OpenAI | GPT-4o, o4 Mini |
| | Anthropic Claude | Claude Sonnet 4.6, Claude Opus 4.6 |
| | Google Gemini | Gemini 2.5 Flash, Gemini 2.5 Pro |
| **Domestic (China)** | DeepSeek | DeepSeek V3, DeepSeek R1 |
| | Qwen (通义千问) | Qwen Max, Qwen Plus, Qwen Turbo |
| | ERNIE (文心一言) | ERNIE 4.5, ERNIE 3.5 |
| **Local Deployment** | Ollama | Llama 3.2, Qwen2.5, Mistral (and more) |
| | Custom | Any OpenAI-compatible API |
**Quick model switching:**
```bash
# Set default model
./scienceclaw config set model.provider claude
./scienceclaw config set model.provider qwen
./scienceclaw config set model.provider ollama
```
For detailed configuration, see the [Configuration Guide](docs/getting-started/configuration.md).
### Add a messaging channel (optional)
```bash
./scienceclaw add telegram # or discord, slack, whatsapp, feishu, matrix
./scienceclaw channels # see what's configured
```
For one-shot mode, skip the TUI entirely:
```bash
./scienceclaw ask "Search TREM2 in Alzheimer's disease and summarize recent findings"
```
---
## What It Can Do
| Capability | Details |
|------------|---------|
| **Search literature** | 15+ sources — PubMed, Semantic Scholar, OpenAlex, Europe PMC, and more |
| **Query databases** | 77+ databases — UniProt, PDB, NCBI, ChEMBL, STRING, GTEx, ClinicalTrials.gov, and more |
| **Run code** | Python, R, Julia via bash — install packages on the fly |
| **Generate figures** | Journal-spec palettes (NPG, Lancet, JCO, NEJM), publication-ready sizing |
| **Write reports** | Real citations from search results, never fabricated |
| **Review research** | 8-dimension ScholarEval rubric for systematic quality assessment |
---
## Channel Integrations
ScienceClaw inherits all channel integrations from OpenClaw. Connect your preferred interface:
| Channel | How to use | Status |
|---------|-----------|--------|
| **Terminal UI** | `scienceclaw tui` | Pre-configured |
| **Web Dashboard** | `scienceclaw dashboard` | Pre-configured |
| **Telegram** | `scienceclaw add telegram` | [Setup guide](docs/channels/telegram.md) |
| **Discord** | `scienceclaw add discord` | [Setup guide](docs/channels/discord.md) |
| **Slack** | `scienceclaw add slack` | [Setup guide](docs/channels/slack.md) |
| **Feishu / Lark** | `scienceclaw add feishu` | [Setup guide](docs/channels/feishu.md) |
| **WhatsApp** | `scienceclaw add whatsapp` | [Setup guide](docs/channels/whatsapp.md) |
| **Matrix** | `scienceclaw add matrix` | [Setup guide](docs/channels/matrix.md) |
| + more | `scienceclaw channels --help` | — |
> **Note:** Feishu is disabled by default. Enable it via `./scienceclaw add feishu` command.
---
## Local & Custom Model Deployment
ScienceClaw supports local model deployment via Ollama or any OpenAI-compatible API. This is ideal for:
- **Privacy-sensitive research** — data never leaves your local machine
- **Cost optimization** — run unlimited queries with local compute
- **Offline usage** — work without internet connectivity
- **Custom models** — deploy specialized models for your domain
### Quick Setup with Ollama
```bash
# 1. Install Ollama
curl -fsSL https://ollama.com/install.sh | sh
# 2. Pull a model
ollama pull qwen2.5:7b
# 3. Add to ScienceClaw
./scienceclaw add model
# Select "ollama" from the preset menu
```
### Using Custom / Private Endpoints
For enterprise deployments or private API servers:
```bash
./scienceclaw add model
# Select "custom" from the menu
# Enter your base URL (e.g., http://your-server:8000/v1)
# Enter your API key and model ID
```
See the [Configuration Guide](docs/getting-started/configuration.md) for detailed setup instructions.
---
## Architecture
```
ScienceClaw = OpenClaw + SCIENCE.md + 264 Skills
```
No TypeScript. No Python servers. No MCP. No plugins. The model does the work.
| Layer | Components |
|-------|-----------|
| **User** | Terminal UI, Web Dashboard, Telegram, Discord, Slack, Feishu, WhatsApp, Matrix |
| **Gateway** | OpenClaw gateway — routes messages, manages sessions, handles tool calls |
| **Agent** | Single `ScienceClaw` agent powered by `SCIENCE.md` (~200 lines) + 264 domain skills |
| **Infrastructure** | `web_search`, `web_fetch`, `bash` — the three built-in OpenClaw tools that do everything |
---
## 🔍 Deep Research
ScienceClaw searches across 15+ sources, cross-references results, and verifies citations before including them in reports.
| Category | Sources |
|----------|---------|
| **Biomedical literature** | PubMed, PubMed Central, Europe PMC |
| **Broad academic** | Semantic Scholar, OpenAlex, CrossRef, CORE |
| **Preprints** | bioRxiv, medRxiv, arXiv |
| **Clinical** | ClinicalTrials.gov, WHO ICTRP |
| **Patents & grants** | Google Patents, NIH RePORTER |
| **General** | Google Scholar, Web search |
---
## 🗄 Database Access
77+ databases across 9 disciplines, all accessed through their public APIs via `web_fetch`.
| Discipline | Databases | Count |
|------------|-----------|-------|
| **Genomics & Transcriptomics** | NCBI Gene, Ensembl, UCSC Genome Browser, GEO, TCGA, GTEx, ENCODE | 10+ |
| **Proteomics & Structure** | UniProt, PDB, AlphaFold DB, InterPro, Pfam, SWISS-MODEL | 8+ |
| **Pathways & Interactions** | STRING, BioGRID, KEGG, Reactome, WikiPathways, IntAct | 8+ |
| **Pharmacology & Drug Discovery** | ChEMBL, DrugBank, PubChem, PharmGKB, DGIdb, TTD | 8+ |
| **Disease & Phenotype** | OMIM, DisGeNET, ClinVar, GWAS Catalog, HPO, Orphanet | 8+ |
| **Immunology** | IEDB, IMGT, ImmPort, TIMER2.0, TCIA | 6+ |
| **Microbiome** | GMrepo, gutMDisorder, BugBase, MicrobiomeDB | 5+ |
| **Clinical & Epidemiology** | ClinicalTrials.gov, GBD, WHO GHO, SEER, cBioPortal | 7+ |
| **Model Organisms** | MGI, FlyBase, WormBase, ZFIN, RGD, SGD | 7+ |
---
## 📚 Skills
264 domain skills provide detailed guidance for specific techniques. Each skill is a markdown file that teaches the model *how* to perform a particular analysis.
| Domain | Example Skills |
|--------|---------------|
| **Bioinformatics** | Differential expression, gene set enrichment, pathway analysis, network construction |
| **Single-cell** | Clustering, trajectory inference, cell type annotation, RNA velocity |
| **Survival analysis** | Kaplan-Meier curves, Cox regression, forest plots, nomograms |
| **Visualization** | Volcano plots, heatmaps, Manhattan plots, Circos plots, UMAP/tSNE |
| **Drug discovery** | Target identification, molecular docking, ADMET prediction, drug repurposing |
| **Clinical** | Meta-analysis, diagnostic test evaluation, risk factor analysis, Mendelian randomization |
| **Genomics** | Variant annotation, GWAS analysis, copy number variation, mutation signatures |
| **Immunology** | Immune infiltration, neoantigen prediction, TCR/BCR repertoire analysis |
| **Machine learning** | Feature selection, model training, cross-validation, SHAP interpretation |
---
## Deployment
### Local (recommended for development)
Already covered in [Quick Start](#quick-start).
### Docker (sandbox only)
The Docker setup provides a sandboxed Python/R/Julia execution environment:
```bash
docker-compose -f docker/docker-compose.yml up
```
See [Deployment Guide](docs/guides/deployment.md) for full production deployment options.
---
## Contributing
Contributions are welcome. Please read [CONTRIBUTING.md](CONTRIBUTING.md) before submitting a pull request.
---
## Author
### Contributors (v1.1.0)
**Li XiaoCong**
Free developer
Contact: [lixy2017@aliyun.com](mailto:lixy2017@aliyun.com)
---
## License
This project is licensed under the [MIT License](LICENSE).
---
**ScienceClaw** — Your AI Research Colleague.
