Jhtdb MCP Server

@pipeworx/jhtdb

Point queries (velocity, pressure, velocity gradient) on Johns Hopkins Turbulence Databases (JHTDB) reference DNS/LES datasets, plus a catalog of every dataset's spatial domain, grid resolution, time range, and available physical fields.

Part of Pipeworx — an MCP gateway connecting AI agents to 1865+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

• jhtdb_datasets(...) — the full catalog of JHTDB datasets (forced isotropic turbulence, channel flow, MHD turbulence, transitional boundary layer, atmospheric boundary layer and wind-farm LES, and more): spatial domain, grid resolution, time range, and available fields for each.
• jhtdb_point_query(...) — interpolated velocity, pressure, or velocity-gradient values at up to 10 points and one time in a named dataset.

Auth

Keyless from the caller's point of view. This pack uses JHTDB's own public testing authorization token internally (documented by the publisher as valid for small queries), and caps every call at 10 points — well under the publisher's documented small-query allowance — so Pipeworx traffic does not load that shared testing quota. There is no BYO-key override: a full (non-testing) JHTDB token is an institutional request against the publisher and is out of scope for this pack unless demand for larger queries shows up.

Data sources

• Johns Hopkins University (Institute for Data Intensive Engineering and Science), Turbulence Databases (JHTDB) — direct numerical simulation (DNS) and large-eddy simulation (LES) of canonical turbulent flows (forced isotropic turbulence at several grid resolutions, channel flow, MHD turbulence, a homogeneous buoyancy-driven flow, a transitional boundary layer, rotating stratified turbulence, stably stratified atmospheric boundary layers, and wind-farm LES), queryable at arbitrary points and times via the publisher's own web service.

The publisher's query interface is SOAP 1.1, not REST/JSON — there is no JSON point-query endpoint to fall back to. The SOAP envelope is built and parsed directly in src/index.ts (regex-based XML matching, consistent with how this codebase already handles SOAP/XML upstreams with no DOMParser available in the Workers runtime — see mcps/sdmx-structure).

The dataset catalog (domain extents, grid resolution, time range, available fields) is transcribed from the publisher's own machine-readable dataset configuration, not derived by probing the service — it is 17 small records, not a generated corpus, so it lives as a plain array in src/index.ts rather than going through the KV baked-index path.

Values are returned in each dataset's native simulation units: dimensionless DNS code units for the periodic-box datasets (isotropic turbulence, channel flow, MHD turbulence), physical SI units (meters, seconds, Pascals) for the atmospheric-boundary-layer and wind-farm LES datasets. jhtdb_point_query's response states which applies for the field just queried.

Dataset codes, enum values for spatial/temporal interpolation methods, and which physical fields each dataset carries are all exactly as the publisher's own API defines them — an interpolation-method typo (e.g. a plausible-looking FD4NoInt instead of the real None_Fd4) fails loudly with the publisher's own validation message rather than silently, which is how the correct enum values here were confirmed.

Quick Start

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

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

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/jhtdb/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 1865+ 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/jhtdb_datasets \
  -H 'Content-Type: application/json' \
  -d '{}'

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

Run it from npm

@pipeworx/mcp-jhtdb 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/jhtdb/mcp endpoint shown above. It needs network access; there is no offline mode.

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

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 Jhtdb data" })

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

More

• Docs and guides
• pipeworx.io

License

MIT