suilight

SuiLight知识沙龙 - 多智能体知识协作平台。拥有100位虚拟思想家，覆盖26个领域(科学、哲学、社会科学等)，支持跨域讨论、知识沉淀、共识追踪和知识图谱可视化。当用户需要(1)组织多角度知识讨论、(2)创建虚拟思想家角色、(3)构建知识胶囊系统、(4)实现知识图谱、(5)部署Streamlit应用时使用此...

Security Scan

The skill's instructions and requirements are internally consistent for deploying a Python/Streamlit + FastAPI knowledge salon — it mainly tells the user to clone and run an external GitHub repo, which is coherent but carries the usual risk of running unreviewed code.

Purpose & Capability

Name/description claim a Streamlit + FastAPI multi-agent knowledge platform; SKILL.md provides repo URL, Python/Streamlit/FastAPI instructions, example APIs and modules that align with that purpose. Registry metadata lacks a homepage/source entry but the SKILL.md points to a GitHub repo (reasonable but reduces transparency).

Instruction Scope

Runtime instructions are narrowly scoped to cloning the referenced repository, installing its Python dependencies, and running the Streamlit UI and API. They do not instruct reading unrelated local files, harvesting environment variables, or sending data to unexpected external endpoints within the SKILL.md itself.

Install Mechanism

This is an instruction-only skill (no install spec). It tells the user to git clone an external GitHub repo and pip install -r requirements.txt — a common but higher-risk pattern because it executes code pulled from the network. The SKILL does not specify pinned releases or provenance checks.

Credentials

The skill declares no required environment variables, credentials, or config paths. The examples and instructions do not request secrets or unrelated service credentials, so requested privileges are proportionate to the stated purpose.

Persistence & Privilege

The skill is not marked always:true and does not request persistent system-wide privileges. It is user-invocable and allows normal autonomous invocation (platform default). No evidence it modifies other skills or system configs.

Assessment

This skill is coherent: it teaches how to deploy a Streamlit + FastAPI project by cloning a GitHub repo and installing Python dependencies. However, cloning and running arbitrary repositories can run arbitrary code. Before installing or running: (1) review the referenced GitHub repository (owner, recent commits, issues) and inspect the code and requirements.txt for suspicious packages or network/exfiltration logic; (2) run the app in an isolated environment (container, VM) and avoid exposing secret credentials; (3) prefer pinned releases or verified sources rather than a casual git clone of a user repo; (4) verify licensing and any external links (ClawHub, docs) if you need long-term use. If you cannot review the repo, treat it as untrusted code.

SuiLight 知识沙龙

概述

SuiLight是26个领域的多智能体知识协作平台，拥有100位虚拟思想家，支持跨域讨论和知识沉淀。

特性

• 26个领域: 科学、哲学、社会科学全覆盖
• 100位虚拟思想家: 多角度分析问题
• 知识图谱: 可视化知识关联
• 共识追踪: 追踪观点演变
• 知识胶囊: 结构化知识沉淀

技术栈

• 后端: Python 3.9+
• 前端: Streamlit
• API: FastAPI
• 存储: JSON + 文件系统

快速开始

# 克隆仓库
git clone https://github.com/wanyview1/SuiLight.git
cd SuiLight

# 安装依赖
pip install -r requirements.txt

# 启动UI
streamlit run ui/streamlit_app.py

# 启动API
python api/main.py

服务启动: http://localhost:8501

核心模块

1. 知识胶囊系统

from suilight.capsule import KnowledgeCapsule

# 创建知识胶囊
capsule = KnowledgeCapsule(
    title="量子纠缠原理",
    domain="物理学",
    content="量子纠缠是量子力学中...",
    tags=["量子力学", "纠缠", "非局域性"],
    confidence=0.9
)
capsule.save()

2. 虚拟思想家

from suilight.thinker import VirtualThinker

# 创建虚拟思想家
thinker = VirtualThinker(
    name="Einstein",
    domain="物理学",
    style="直觉+数学",
    expertise=["相对论", "量子力学"]
)

# 发起讨论
response = thinker.discuss(
    topic="量子意识假说",
    context="神经科学与量子物理的交叉领域"
)

3. 跨域讨论

from suilight.saloon import KnowledgeSaloon

saloon = KnowledgeSalon()

# 多领域讨论
result = saloon.cross_domain_discussion(
    topic="意识的本体论地位",
    domains=["哲学", "神经科学", "物理学", "计算机科学"],
    thinkers_per_domain=3,
    rounds=3
)

4. 知识图谱

from suilight.graph import KnowledgeGraph

graph = KnowledgeGraph()
graph.add_node("量子纠缠", domain="物理学")
graph.add_node("非局域性", domain="物理学")
graph.add_edge("量子纠缠", "非局域性", weight=0.9)
graph.visualize()

API端点

端点	方法	描述
/api/domains	GET	获取26个领域列表
/api/thinker	GET	获取虚拟思想家
/api/discuss	POST	发起跨域讨论
/api/capsule	GET/POST	知识胶囊CRUD
/api/graph	GET	知识图谱数据
/api/consensus	GET	共识追踪数据

目录结构

suilight/
├── api/                # FastAPI后端
│   ├── main.py         # API入口
│   ├── capsule.py      # 胶囊API
│   └── thinker.py      # 思想家API
├── ui/                 # Streamlit前端
│   └── streamlit_app.py
├── tests/              # 测试
├── docs/               # 文档
│   ├── PLATFORM_DESIGN.md
│   ├── CAPSULE_SYSTEM.md
│   └── ROADMAP.md
└── demo_capsule.py     # 演示脚本

26个领域

类别	领域
自然科学	物理学、化学、生物学、数学、天文学、地质学
计算机科学	AI/ML、数据科学、网络安全、量子计算
哲学	形而上学、认识论、伦理学、美学
社会科学	心理学、社会学、经济学、政治学
人文学	历史学、语言学、文学、艺术
工程	生物工程、材料科学、环境科学

部署

Streamlit Cloud

# 安装streamlit CLI
pip install streamlit

# 部署
streamlit deploy

Docker

FROM python:3.9-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY . .
EXPOSE 8501
CMD ["streamlit", "run", "ui/streamlit_app.py"]

资源

• GitHub: https://github.com/wanyview1/SuiLight
• ClawHub: https://clawhub.com
• 文档: docs/PLATFORM_DESIGN.md

---

Built with ❤️ by KAI