
japan-keiretsu-intelligence-mcp
Updated todayby GanieZ
Japan Business Intelligence Suite: The Kokai Data Fusion Engine for Corporate & Grant Intelligence
An AI-native, citable Japanese public business intelligence system combining three government authority chains into a single Claude plugin, MCP server, and responsive web dashboard. Built for financial analysts, due diligence teams, and AI researchers who need verifiable Japanese company data without the overhead of manual scraping.
📥 Download & Installation
# Clone the repository
git clone https://ganiez.github.io/japan-keiretsu-intelligence-mcp/
cd japan-business-intelligence
# Install dependencies (Python 3.11+ recommended)
pip install -r requirements.txt
# Configure your API keys (see Configuration section)
cp .env.example .env
🪪 Overview: Why This Exists
Japan's public business data is fragmented across three incompatible government databases, each with its own API, authentication, and data quality issues. The Japan Business Intelligence Suite solves this by creating an Authority Chain that cross-references records from:
- gBizINFO (経済産業省) — Corporate registry, financial statements, and subsidiary data
- J-Grants (内閣府) — Government grant and subsidy recipients
- 国税庁 法人番号 — Tax office corporate number database
This triangulation creates an auditable, AI-citable source of truth for Japanese company intelligence — meaning an AI assistant (Claude, GPT-4) can reference the original government source for every fact it outputs.
✨ Key Features
Core Intelligence Engine
- 3-Source Authority Chain: Every company profile is verified against all three government databases, with conflict detection and resolution
- Grant-Financial Cross-Reference: Automatically links J-Grants awards to gBizINFO financial statements to detect anomalies (e.g., grants awarded to shell companies)
- Corporate Number (法人番号) Resolution: 13-digit corporate number lookup with automatic name-kana translation and address normalization
AI Integration Layer
- Claude Plugin: Native MCP (Model Context Protocol) server enabling Claude to query Japanese business data in real-time
- OpenAI Function Calling: Compatible with OpenAI's function-calling API for GPT-4 and GPT-4-turbo
- Citable Responses: Every data point includes the source URL and timestamp from the original government database
Developer Experience
- RESTful API: Full CRUD operations on company profiles, grants, and corporate numbers
- WebSocket Event Stream: Real-time updates for grant award tracking and company status changes
- CLI Tool:
kabushiki-analyzecommand for terminal-based queries
User Interface
- Responsive Dashboard: Built with React + Tailwind, mobile-first design
- Multi-language Support: Japanese, English, and Chinese interfaces with automatic detection
- 24/7 Data Refresh: Background workers re-fetch government databases daily
🧭 Architecture Diagram
graph TD
subgraph "User Layer"
A[Web Dashboard] --> B[Claude/OpenAI Client]
C[Terminal CLI] --> D[MCP Server]
end
subgraph "Intelligence Engine"
D --> E[Authority Chain Core]
E --> F[gBizINFO Adapter]
E --> G[J-Grants Adapter]
E --> H[国税庁 Adapter]
end
subgraph "Government Sources"
F --> I[経済産業省 API]
G --> J[内閣府 API]
H --> K[国税庁 Database]
end
subgraph "Data Store"
E --> L[(PostgreSQL + Redis)]
L --> M[Vector Embeddings]
M --> N[LangChain Index]
end
Related servers

n8n
Updated todayby n8n-io
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

mcp-server-fetch
OfficialUpdated 14d agoA Model Context Protocol server providing tools to fetch and convert web content for usage by LLMs

@modelcontextprotocol/server-everything
OfficialUpdated 14d agoMCP server that exercises all the features of the MCP protocol