mcpField
I
Ioos Erddap

Ioos Erddap

Remote
pipeworx.ioScience & Education0Updated 2d ago

About Ioos Erddap

An MCP server for US ocean-observing data from IOOS regional ERDDAP services, aimed at ocean and environmental data analysts. It supports dataset discovery, station and glider time series, and point queries on model or radar grids.

Key features

  1. 01Search regional dataset catalogs
  2. 02Inspect variables, units, and coverage
  3. 03Read station and glider time series
  4. 04Retrieve values at grid locations

Use cases

  1. 01Find ocean sensor observations
  2. 02Analyze glider deployment time series
  3. 03Query regional model data at a depth

MCP support

It exposes dataset search, dataset metadata, tabular time-series, and gridded point-query tools to MCP clients.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://gateway.pipeworx.io/ioos-erddap/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

@pipeworx/ioos-erddap

US Integrated Ocean Observing System data — buoys, tide gauges, gliders, HF radar currents, water quality and regional ocean model grids — read live from the IOOS regional associations' ERDDAP servers.

Part of Pipeworx — an MCP gateway connecting AI agents to 1881+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

  • ioos_erddap_search_datasets(query, protocol?, node?, limit?) — full-text search of one node's catalogue. Start here: every other tool takes a dataset_id from these results.
  • ioos_erddap_dataset_info(dataset_id, node?) — variables, units, axes, time coverage, licence.
  • ioos_erddap_tabledap(dataset_id, variables?, constraints?, node?, limit?) — station and glider timeseries rows.
  • ioos_erddap_griddap_point(dataset_id, variable, latitude, longitude, time?, depth?, node?) — one gridded value from a regional model or radar grid.

Auth

Keyless. Every node serves anonymously.

Data sources

Six nodes, selected with node. A dataset lives on exactly one of them, so getting node right is the first thing a caller has to do; a wrong id raises an error that lists them all.

nodeServerCoverage
nationalhttps://erddap.ioos.us/erddapIOOS-wide inventories, asset and metric datasets
sensors (default)https://erddap.sensors.ioos.us/erddapthe national in-situ sensor network
glidershttps://gliders.ioos.us/erddapevery US underwater glider deployment
paciooshttps://pae-paha.pacioos.hawaii.edu/erddapHawaii and the Pacific Islands; ROMS and WRF grids
secoorahttps://erddap.secoora.org/erddapthe US Southeast
gcooshttps://erddap.gcoos.org/erddapthe Gulf of Mexico

NERACOOS and NANOOS are deliberately absent. https://www.neracoos.org/erddap and https://data.nanoos.org/erddap both answer 301 from our egress (measured 2026-09-17). Adding them without resolving where they redirect would ship two node ids that fail for every caller. If you add them, verify a live search first, not just a reachable root.

Traps

Protocol-level traps live in shared/src/erddap.ts (shared with noaa-coastwatch) and are documented in full at the top of that file. Two matter most here:

  • erddap.ioos.us has almost no griddap. The national node is inventories and metrics; the gridded fields are on the regional nodes. A griddap_point call against national is not a bug.
  • PacIOOS ROMS grids are 4-D: time, depth, latitude, longitude. depth is required on those, and omitting it is an axis error rather than a defaulted surface reading. depth: 0.25 is the shallowest level of roms_hiig.

And the one this pack found the hard way: ERDDAP's first &-segment is positional — the server reads it as the variable list whatever it contains. Asking for every column (no variables) while dropping the now-empty first segment slides the row-cap expression into the variable slot and the server answers Unrecognized variable="orderByLimit("5")", which reads as a broken dataset rather than a malformed URL. The shared client keeps the empty segment.

Quick Start

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

{
  "mcpServers": {
    "ioos-erddap": {
      "url": "https://gateway.pipeworx.io/ioos-erddap/mcp"
    }
  }
}
What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/ioos-erddap/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 1881+ 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/ioos_erddap_search_datasets \
  -H 'Content-Type: application/json' \
  -d '{"query":"water temperature","node":"sensors","limit":5}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/ioos_erddap_search_datasets. 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:

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

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-ioos-erddap

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 Ioos Erddap data" })

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

More

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.