Fujitsu Social Digital Twin MCP Server

This integration bridges Fujitsu's core Digital Twin and Rehearsal APIs with the Model Context Protocol (MCP), enabling Large Language Models (LLMs) to command Fujitsu's Digital Rehearsal capabilities via natural language.

System Summary

The Fujitsu Social Digital Twin system models the digital state of populations and physical assets based on real-world inputs, extending to encompass entire economic and societal frameworks. Its centerpiece, the 'Digital Rehearsal' function, permits the pre-verification of proposed interventions by simulating human and societal responses in a virtual environment, assessing potential impacts before physical deployment.

This MCP wrapper establishes connectivity between LLMs and the Digital Rehearsal API, translating user intent expressed conversationally into actionable simulation execution and result interpretation.

Core Functionality

• Inventory and visualization of extant simulation runs
• Initiation of new simulation executions
• Extraction and in-depth examination of simulation outputs
• Management and retrieval of simulation artifacts
• Specialized analysis of traffic dynamics simulations
• Comparative study across distinct scenario configurations
• Automated generation of simulation parameters from descriptive text

Operational Requirements

• Execution environment running Python version 3.13 or newer
• Valid credentials for the Fujitsu API Gateway (API Key required)
• An MCP-compliant LLM interface (e.g., Claude Desktop client)

Deployment Instructions

Installation via Smithery

For automated deployment to Claude Desktop using Smithery:

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

1. Source Code Acquisition

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

2. Environment Preparation

Using uv (Recommended Method):

First, install the uv tool:

# Standard pip installation
pip install uv

# Or via curl (for Unix-like systems)
curl -sSf https://astral.sh/uv/install.sh | sh

Then, configure the environment using uv:

# Create an isolated virtual environment
uv venv

# Activation commands:
# On Windows:
.venv\Scripts\activate
# On Unix/macOS:
source .venv/bin/activate

# Install all necessary dependencies
uv pip install -r requirements.txt

Alternatively, utilize the provided setup script:

# Grant execution permission
chmod +x setup.sh

# Execute setup routine
./setup.sh

3. Credential Configuration

Specify necessary access tokens and endpoints via environment variables:

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

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

Alternatively, populate a .env file located in the root directory:

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

Operation Guide

Launching the MCP Service

python -m fujitsu_sdt_mcp

The server interfaces with MCP clients via standard input/output streams.

Direct Client Interaction

A minimal command-line client is available for direct testing:

python client.py

Upon launch, this client displays all accessible resources and tools, allowing for direct invocation from the terminal.

Integration with Claude Desktop

To deploy this server alongside MCP-enabled assistants like Claude Desktop:

1. Open the Claude Desktop application.
2. Navigate to Settings > Developer Settings > Edit Config File.
3. Insert the following configuration block:

"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"
}

1. Relaunch Claude Desktop for changes to take effect.

Available Interfaces and Utilities

Resources (Informational Endpoints)

• resource://digital_rehearsal_overview: Technical summary of the Digital Rehearsal capability.
• resource://simulation_metrics_explanation: Glossary for simulation measurement criteria.
• resource://scenario_examples: Curated collection of sample simulation blueprints.

Tools (Actionable Functions)

• list_simulations: Query for the catalog of stored simulations.
• start_simulation: Initiate a new rehearsal process.
• get_simulation_result: Fetch and present the outcomes of a completed simulation.
• get_metrics: Retrieve quantitative performance indicators from a run.
• list_simdata: Obtain a manifest of available input datasets.
• get_simdata: Access specific input datasets.
• analyze_traffic_simulation: Perform detailed post-mortem analysis on traffic models.
• compare_scenarios: Execute side-by-side comparison logic between two or more configurations.
• create_natural_language_simulation_config: Translate descriptive text directly into executable simulation parameters.

Operational Examples

Querying Simulation List

tool list_simulations

Launching a Rehearsal

tool start_simulation {"simdataId": "your_simdata_id"}

Traffic Model Scrutiny

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

Auto-Configuration from Text

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

Illustrative LLM Dialogue Flow

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?

Collaboration and Support

Bugs or feature suggestions should be submitted via GitHub Issues. Contributions via Pull Requests are encouraged.

Licensing

Distributed under the terms of the MIT License. Consult the LICENSE file for specifics.

Acknowledgements

• Fujitsu Corporation - For pioneering the Social Digital Twin framework.
• Model Context Protocol (MCP) Community - For enabling robust LLM integration.

WIKIPEDIA: Business management tools encompass all methodologies, applications, controls, computational solutions, and systems utilized by enterprises to navigate market shifts, maintain competitive parity, and enhance operational efficiency. These tools span functional areas such as strategic planning, process oversight, record-keeping, personnel administration, decision support, and performance monitoring.

== System Classification == Tools can be segmented based on organizational department or specific management aspect (e.g., planning, control, process automation). A functional breakdown often includes:

Tools for enterprise-wide data ingestion and verification. Utilities designed for business process governance and refinement. Platforms for data aggregation and strategic decision synthesis. The evolution of these managerial instruments has been explosive over the last decade due to rapid technological advancement, making the selection process complex. This challenge is driven by constant pressures to reduce expenditure, maximize revenue streams, deeply understand client requirements, and deliver offerings precisely as demanded. In this environment, leadership must adopt a strategic posture toward tool adoption, rather than merely chasing the newest release. Relying on off-the-shelf tools without tailoring them often leads to systemic instability; therefore, business management utilities must be chosen deliberately and then tailored to the organization's specific operational mandates.

== Global Adoption Trends (2013 Survey Highlight) == A 2013 survey by Bain & Company quantified the worldwide utilization of business instruments, reflecting how their output addresses regional imperatives amidst varying market conditions. The top ten instrumental categories identified were:

Strategic planning frameworks Client relationship management systems Personnel sentiment surveys Competitive benchmarking practices Balanced scorecard methodologies Core competency identification Outsourcing governance Organizational change management programs Supply chain optimization Mission/Vision statement formalization Market segmentation analysis Total Quality Management (TQM)

== Commercial Software Definition == Software solutions or integrated suites deployed by business users to execute various operational tasks are termed business applications. These applications are engineered to augment productivity, accurately measure performance metrics, and execute diverse organizational functions with precision. The trajectory began with Management Information Systems (MIS), expanded into Enterprise Resource Planning (ERP), subsequently incorporated Customer Relationship Management (CRM), and has largely transitioned into the realm of cloud-based business management services. While a clear relationship exists between IT investments and organizational effectiveness, the true value amplification stems from two key factors: the efficacy of the deployment process and the rigor applied during the selection and adaptation of the requisite tools.

== Tools for Small and Medium Enterprises (SMEs) == SME-focused tools are crucial as they furnish mechanisms designed to conserve resources...