Skip to main content
Glama
jp-nite-accidents

jp-nite-accidents

by pipeworx-io

@pipeworx/jp-nite-accidents

Japan's product accident reports as the National Institute of Technology and Evaluation (NITE) publishes them in SAFE-Lite: accidents notified under the Consumer Product Safety Act, major and non-major, with the product, model, the business, the accident as notified, the cause NITE found and the recurrence measures taken (including recalls and exchanges).

Part of Pipeworx — an MCP gateway connecting AI agents to 2016+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

  • jp_nite_accidents_search(query?, severity?, date_from?, date_to?, page?, limit?) — free-word search over the reports, newest accident date first (an empty query lists the newest). Each hit: report number, major or non-major, accident date, product, model, business, the accident as notified with the prefecture, cause category and the record link. severity is all, major (重大製品事故: death, serious injury, fire) or non_major. Search Japanese words (充電器, 石油ストーブ, a maker's name); full and half width are folded. date_from/date_to filter the one page read (100 rows) and the answer states whether the window was fully covered; page fetches the next page. Zero matches states the list and severity searched.

  • jp_nite_accident_get(accident_number) — one report (such as A202500845 for a major accident or 2025-1924 for a non-major one): harm, usage period, the cause finding, cause category, recurrence-prevention measures, who notified it and when it was received. An unknown number is found:false.

Related MCP server: deeprecall-mcp

Auth

Keyless.

Data sources

  • NITE (製品評価技術基盤機構), product accident database SAFE-Lite (製品事故情報): read live per call, one request per search and two per record (a number lookup, then the record). Nothing is stored. Reports appear once their investigation is finished, so the newest accidents lag by months. A day shown as 00 in the source is kept as a month (2025-11), never as a made-up day.

Traps worth knowing: a report's own record page is keyed by an internal id, so jp_nite_accident_get first finds the id through the report number. The cause category is blank on many major accidents ("not identified"). Searches are Japanese: an English query finds only Latin text in model names. The recall notices that Japan's Consumer Affairs Agency publishes are a different dataset (jp-recalls); this one is the accident record, where a recall appears only as a recurrence measure.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "jp-nite-accidents": {
      "url": "https://gateway.pipeworx.io/jp-nite-accidents/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/jp-nite-accidents/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 2016+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/jp_nite_accidents_search \
  -H 'Content-Type: application/json' \
  -d '{"query":"リチウム電池 充電","severity":"major","limit":3}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/jp_nite_accidents_search. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Run it from npm

@pipeworx/mcp-jp-nite-accidents is a local stdio bridge to the Pipeworx gateway: your MCP client starts it with npx, and it forwards every call to the same https://gateway.pipeworx.io/jp-nite-accidents/mcp endpoint shown above. It needs network access; there is no offline mode.

{
  "mcpServers": {
    "jp-nite-accidents": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-jp-nite-accidents"]
    }
  }
}

Without a key it calls anonymously, capped at 50 calls a day. A free account raises that to 200 a day: get a key at https://pipeworx.io/signup and add "env": { "PIPEWORX_API_KEY": "<your key>" } to the server entry. The key is sent as a Bearer token, the same way a hosted connection sends it, so calls are metered and billed exactly as they are there.

It lists only this pack's tools by default. Set PIPEWORX_META_TOOLS=1 to also list ask_pipeworx and the other gateway-wide tools.

Behind an HTTPS proxy: Node's built-in fetch does not read HTTPS_PROXY by default. Newer Node releases honour it when NODE_USE_ENV_PROXY=1 is set; check your Node version's documentation.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Jp Nite Accidents data" })

The gateway picks the right tool and fills the arguments automatically.

More

License

MIT

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI agents to search de-identified NASA aviation safety incident reports by keyword, date range, aircraft or flight phase, fetch a full narrative and coded fields by ACN, and list the thirty covered safety topics with their report counts and NASA update dates. Answers come from NASA's periodic Report Set PDFs parsed into a queryable store, so coverage is limited to those topics.
      93 npm
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables agents to check whether a Japanese vehicle maker or importer has filed a recall or improvement measure with Japan's Ministry of Land, Infrastructure, Transport and Tourism, filtering the ministry's public notification lists by company name, kind, and year. It also retrieves individual press releases with model names, filing dates, full legal company names, and attached-document details.
      MIT