mcp-world-port-index

# @pipeworx/world-port-index

Port facilities, depths, anchorages and navigational detail for ~2,950 ports
worldwide — lookup by name/UN-LOCODE/WPI number, radius search from a lat/lon
point, and listing by country.

Part of [Pipeworx](https://pipeworx.io) — an MCP gateway connecting AI agents to 1808+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

## Tools

- `port_lookup(name | unlocode | wpi_number)` — resolve one port and return
  its WPI number, coordinates, harbor size/type, channel/anchorage/cargo-pier
  depths, and facility flags (pilotage, tugs, drydock, railway). Refuses with
  an error if nothing matches; returns a short candidate list if a name
  matches more than one port.
- `ports_near(latitude, longitude, radius_km?, limit?)` — ports within a
  radius of a point, nearest first, with distance in km.
- `ports_by_country(country, harbor_size?, limit?)` — ports for a country
  (name or ISO 3166-1 alpha-2 code), optionally filtered by harbor size
  (V=Very small, S=Small, M=Medium, L=Large).

## Auth

Keyless. The pack reads its dataset from a Pipeworx-side store (see "Data
sources" below) and makes no outbound calls at request time.

## Data sources

- U.S. National Geospatial-Intelligence Agency (NGA), Maritime Safety
  Information, World Port Index (Pub. 150) — the standard reference for port
  facilities, depths, and navigational characteristics worldwide. US federal
  government data; public domain.

This pack does **not** bake its data into the gateway bundle — a first
version did, and a 1.7MB baked table pushed the gateway's JS bundle over its
hard size ceiling, blocking every lane's deploy (fleet #3061, 2026-10-14).
Instead it reads a single JSON object out of the shared `pipeworx-datasets`
R2 bucket at request time, cached in Worker-isolate memory for several hours
so only a cold isolate pays the fetch. `scripts/refresh-ports.mjs` fetches
NGA's live feed, trims it, and writes that object locally; an operator with
R2 write access runs the `wrangler r2 object put` command it prints to
publish it (never run by this script itself). If the R2 object isn't
published yet, or the binding isn't configured on a deployment, every tool
returns a `available: false` note rather than throwing or going silently
empty. Every real response carries `data_as_of`. Depth fields
(`channelDepth`, `anchorageDepth`, `cargoPierDepth`, `oilTerminalDepth`) are
in meters as published by NGA; `null` means NGA records no data for that
category, not zero depth. `harborTypeCode` and `shelterCode` are passed
through as NGA's own Pub 150 codes rather than decoded, since only the
harbor-size code (V/S/M/L) has an unambiguous, well-documented mapping.

## Quick Start

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

```json
{
  "mcpServers": {
    "world-port-index": {
      "url": "https://gateway.pipeworx.io/world-port-index/mcp"
    }
  }
}
```

### What this endpoint actually serves

`tools/list` at `https://gateway.pipeworx.io/world-port-index/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:

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

Both URLs reach the same gateway and the same 1808+ 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

```bash
curl -X POST https://gateway.pipeworx.io/v1/tools/port_lookup \
  -H 'Content-Type: application/json' \
  -d '{"name":"Rotterdam"}'
```

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

## Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no
gateway round-trip:

```json
{
  "mcpServers": {
    "world-port-index": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-world-port-index"]
    }
  }
}
```

Or run it directly to confirm it starts:

```bash
npx -y @pipeworx/mcp-world-port-index
```

It speaks MCP over stdin/stdout and answers `initialize`/`tools/list`/`tools/call`
for **only** this pack's tools — none of the shared meta-tools the gateway
connection above adds. Same source, same tools, no ask_pipeworx routing.

## 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 World Port Index data" })
```

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

## More

- [Docs and guides](https://pipeworx.io/docs)
- [pipeworx.io](https://pipeworx.io)

## License

MIT