富士通社交数字孪生 MCP 服务器

该项目将富士通的社交数字孪生和数字排练 API与模型上下文协议 (MCP)相集成，允许大型语言模型 (LLM) 通过自然语言访问富士通的数字排练 API。

概述

富士通的“社会数字孪生”不仅能够基于现实世界的数据，在数字空间中重现人和物体的状态，还能重现整个经济和社会活动。其核心功能“数字演练”使用户能够在现实世界采取措施之前，在数字空间中模拟人类和社会行为，从而提前验证其效果和影响。

该项目使用 MCP 来弥合 LLM 和数字排演 API 之间的差距，使用户能够运行模拟并使用自然语言分析结果。

主要特点

检索并显示模拟列表
开始模拟
检索和分析模拟结果
管理模拟数据
分析交通模拟
比较场景
从自然语言生成模拟配置

先决条件

Python 3.13 或更高版本
访问富士通 API 网关（API 密钥）
MCP 兼容的 LLM 客户端（例如 Claude Desktop）

安装

通过 Smithery 安装

要通过Smithery自动为 Claude Desktop 安装 Fujitsu Social Digital Twin Server：

npx -y @smithery/cli install @3a3/fujitsu-sdt-mcp --client claude

1.克隆存储库

git clone https://github.com/3a3/fujitsu-sdt-mcp.git
cd fujitsu-sdt-mcp

2. 设置环境

使用 uv (推荐) ：

首先，安装uv：

# Install uv using pip
pip install uv

# Or using curl (Linux/macOS)
curl -sSf https://astral.sh/uv/install.sh | sh

然后，使用 uv 设置你的环境：

# Create virtual environment
uv venv

# Activate virtual environment
# Windows:
.venv\Scripts\activate
# Unix/MacOS:
source .venv/bin/activate

# Install dependencies
uv pip install -r requirements.txt

或者，您可以使用提供的安装脚本：

# Make the script executable
chmod +x setup.sh

# Run setup
./setup.sh

3.设置环境变量

# Windows:
set FUJITSU_API_BASE_URL=https://apigateway.research.global.fujitsu.com/sdtp
set FUJITSU_API_KEY=your_api_key_here

# Unix/MacOS:
export FUJITSU_API_BASE_URL=https://apigateway.research.global.fujitsu.com/sdtp
export FUJITSU_API_KEY=your_api_key_here

或者，您可以创建具有以下设置的.env文件：

FUJITSU_API_BASE_URL=https://apigateway.research.global.fujitsu.com/sdtp
FUJITSU_API_KEY=your_api_key_here

用法

启动 MCP 服务器

python -m fujitsu_sdt_mcp

服务器使用标准 I/O 与 MCP 客户端进行通信。

使用交互式客户端

还提供了一个简单的客户端，用于与 MCP 服务器直接交互：

python client.py

当您启动客户端时，将显示可用资源和工具的列表，您可以从命令行调用它们。

使用 Claude Desktop 进行设置

要将服务器与 MCP 兼容的 LLM 客户端（例如 Claude Desktop）一起使用：

打开 Claude 桌面
前往“设置”>“开发者设置”>“编辑配置文件”
添加以下设置：

"command": "uvx",
"args": [
    "fujitsu-sdt-mcp"
],
"env": {
    "FUJITSU_API_BASE_URL": "https://apigateway.research.global.fujitsu.com/sdtp",
    "FUJITSU_API_KEY": "your_api_key_here"
}

重启Claude桌面

资源和工具

资源

resource://digital_rehearsal_overview ：数字排练技术概述
resource://simulation_metrics_explanation ：模拟指标的解释
resource://scenario_examples ：示例场景

工具

list_simulations ：检索模拟列表
start_simulation ：开始模拟
get_simulation_result ：检索模拟结果
get_metrics ：检索模拟指标
list_simdata ：检索模拟数据列表
get_simdata ：检索模拟数据
analyze_traffic_simulation ：分析交通模拟
compare_scenarios ：比较场景
create_natural_language_simulation_config ：从自然语言生成模拟设置

使用示例

检索模拟列表

tool list_simulations

开始模拟

tool start_simulation {"simdataId": "your_simdata_id"}

分析交通模拟

tool analyze_traffic_simulation {"simulationId": "your_simulation_id", "region": "Tokyo", "timeRange": "morning rush hour", "scenario": "traffic optimization"}

从自然语言生成模拟设置

tool create_natural_language_simulation_config {"description": "I want to run a traffic optimization simulation during morning rush hour in Tokyo"}

LLM 互动示例

User: I want to run a traffic simulation during morning rush hour in Tokyo

Claude: I can help you run a traffic simulation for Tokyo's morning rush hour. Here's how we can set it up.

First, let's generate simulation settings from your natural language description:
[Tool execution: create_natural_language_simulation_config]

Generated settings:
{
  "simulationType": "traffic",
  "parameters": {
    "region": "Tokyo",
    "timeRange": "morning_rush"
  },
  "name": "traffic_20250416_123045"
}

Next, let's check available simulation data:
[Tool execution: list_simdata]

From this data, I can start a simulation using Tokyo's morning traffic data.
I've selected the appropriate simulation data ID and will start the simulation:
[Tool execution: start_simulation]

The simulation has started. The simulation ID is "sim_12345".
It will take some time for the results to be available. Once complete, I'll analyze them.

(After simulation completes)
Let's analyze the simulation results:
[Tool execution: analyze_traffic_simulation]

Analysis results:
- CO2 emissions: 1,250kg
- Average travel time: 42 minutes
- Total traffic volume: 15,700 vehicles

From these results, I can see that the Shibuya-Shinjuku route experiences particularly severe congestion.
Here are some potential improvements:
1. Optimize traffic signal control
2. Implement one-way traffic on certain roads during specific times
3. Increase public transportation frequency

Would you like to see a more detailed analysis?

贡献

请通过 GitHub Issues 报告错误或功能请求。欢迎提交 Pull 请求。

执照

本项目遵循 MIT 许可证发布。详情请参阅许可证文件。

致谢

富士通公司 - 社交数字孪生技术的发展
模型上下文协议 (MCP) - 实现与 LLM 的无缝集成

You must be authenticated.

security – no known vulnerabilities

license - permissive license

quality - confirmed to work

remote-capable server

The server can be hosted and run remotely because it primarily relies on remote services or has no dependency on the local environment.

Tools

View all tools

使法学硕士能够通过自然语言访问富士通的数字排练 API，从而允许用户在数字空间中运行和分析人类和社会行为的模拟。

Related MCP Servers

applescript-mcp
joshrutkowski
A
security
A
license
A
quality
Enables LLM applications to interact with macOS through AppleScript.
Last updated -
21
138
TypeScript
MIT License
MCP Toolbox
ai-zerolab
-
security
A
license
-
quality
A comprehensive toolkit that enhances LLM capabilities through the Model Context Protocol, allowing LLMs to interact with external services including command-line operations, file management, Figma integration, and audio processing.
Last updated -
17
Python
Apache 2.0
Lightning Tools MCP Server
getAlby
-
security
F
license
-
quality
Enables interaction with lightning addresses and common lightning tools via your LLM, providing Lightning Network functionality through natural language.
Last updated -
10
1
TypeScript
Vibe-Eyes
monteslu
-
security
F
license
-
quality
An MCP server that enables LLMs to 'see' what's happening in browser-based games and applications through vectorized canvas visualization and debug information.
Last updated -
35
23
JavaScript

View all related MCP servers

Fujitsu Social Digital Twin MCP Server