Japan Data MCP

A Model Context Protocol (MCP) server that gives AI agents cross-source access to Japanese public data — law, corporations, and statistics — normalized to English metadata, with the required source attribution on every response.

Hosted and free. No install, no API key for the law tools. Original texts are authentic Japanese from official government sources; English fields are normalized metadata, not unofficial translations.

Connect

Claude Code:

claude mcp add --transport http japan-data https://japan-data-mcp.netlify.app/mcp

Claude Desktop / Cursor / any MCP client (Streamable HTTP):

{
  "mcpServers": {
    "japan-data": {
      "type": "streamable-http",
      "url": "https://japan-data-mcp.netlify.app/mcp"
    }
  }
}

Tools

Tool	What it does	API key
`search_japanese_law`	Find laws by title or full text; relevance-ranked, English metadata	none
`get_law_text`	Fetch the authentic Japanese text of a law, whole or per article	none
`search_corporation`	Look up corporations by name or 13-digit corporate number	free appId
`search_statistics`	Search official statistics tables (e-Stat)	free appId

Typical agent flow: search_japanese_law → get_law_text — identify the law, then quote its exact text. Both steps need no key.

Try it

curl -s https://japan-data-mcp.netlify.app/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{
    "name":"get_law_text",
    "arguments":{"law_id":"321CONSTITUTION","article":"9"}}}'

A plain GET /mcp returns server info and the tool list (liveness check).

Usage transparency

Anonymous aggregate usage is public at /stats. No raw IPs or user agents are stored — clients are counted via salted hashes only.

Stage 0 is a minimal, free deployment to measure real demand. Every response carries the source attribution required by the Government of Japan Standard Terms of Use v2.0. This is an independent data-access utility, not affiliated with the Japanese government. Source: github.com/yurukusa/japan-data-mcp (MIT).