AQUAVIEW MCP

by aquaview-dah

Not rated
GitHub

About

Access Oceanic and Atmospheric Data using AI Agents

Details

Author
aquaview-dah
Categories
Developer Tools, Other

Setup

Install AQUAVIEW MCP in your MCP client (Claude Desktop, Cursor, Windsurf, and others).

Repository: https://github.com/aquaview-dah/mcp

Follow the installation instructions in the repository README, then restart your MCP client.

Query 700,000+ ocean & atmospheric datasets from inside Claude, ChatGPT, Gemini, Cursor, and any MCP client — no API keys, no schemas, no glue code.

Homepage·MCP Overview·Install·Prompts·Examples·Notebooks·Docs

AQUAVIEW MCP is a hostedModel Context Protocolserver that gives any LLM agent direct, structured access to a unified catalog of700,000+ oceanographic, atmospheric, and marine datasetsdrawn from68 authoritative sources— NOAA, all 11 IOOS Regional Associations, the World Ocean Database, the Global Argo array, GOES-R satellites, NEXRAD weather radar, ESA Sentinel, IFREMER, EMODnet, Ocean Networks Canada, and more.

Ask in plain English. AQUAVIEW finds, filters, aggregates, and returns the data — including direct download links to NetCDF, GRIB2, GeoTIFF, and CSV files where they exist.

"Find me all NDBC buoys in the Gulf of Mexico that recorded wave heights over 6 meters during Hurricane Ian in September 2022, and give me their data files."

The model uses AQUAVIEW's tools to scope by collection, bounding box, and time, apply CQL2 filters on per-variable statistics, and stream back STAC items with assets.

The AQUAVIEW MCP server is hosted athttps://mcp.aquaview.org/mcp(HTTP transport). No installation. No keys.

claude mcp add --transport http aquaview https://mcp.aquaview.org/mcp

Settings → Developer → Edit Configand merge:

{ "mcpServers": { "aquaview": { "type": "http", "url": "https://mcp.aquaview.org/mcp" } } }

Settings → MCP → Add new MCP Server, paste the URL above, typehttp.

ChatGPT, Gemini, OpenAI Agents SDK, Anthropic API direct, and more

SeeINSTALL.mdfor every supported client.

Drop these into any MCP-enabled chat to see AQUAVIEW in action:

68 collections, grouped. Full table with bbox/temporal coverage and variable lists indocs/collections.md.

AQUAVIEW exposes four MCP tools. Full schemas indocs/tools-reference.md.

The catalog is built on theSTACspecification, so items, properties, and assets follow a stable, well-documented schema.

Twenty-one walkthroughs inexamples/, each a self-contained scenario with the prompt, a real transcript, the result, and variations to try.

- 02-sea-surface-temperature/— SST near a region using CoastWatch + WOD + buoys
-
03-hurricane-tracking/— track + intensity from AOML, GOES-R, NDBC, GDP
-
05-argo-float-profiles/— temperature/salinity profiles from the global Argo array

Three Jupyter notebooks innotebooks/showing AQUAVIEW MCP through the native agent connectors of each major LLM provider:

- 01-claude-mcp-agent.ipynb— Anthropic Python SDK withmcp_serversparameter
- 02-chatgpt-mcp-agent.ipynb— OpenAI Agents SDK with hosted MCP tool
- 03-gemini-mcp-agent.ipynb— Google Gemini Agent SDK with MCP tool

Each notebook asks the same research question (Argo float profiles around Hawaii) so you can compare how each agent reasons over the same catalog.

┌─────────────────────────┐ ┌──────────────────────────┐ │ Claude / ChatGPT / │ │ AQUAVIEW MCP Server │ │ Gemini / Cursor / │ ──HTTP─▶│ mcp.aquaview.org/mcp │ │ any MCP client │ │ │ └─────────────────────────┘ │ list_collections │ │ search_datasets │ │ aggregate │ │ get_item │ └────────────┬─────────────┘ │ ▼ ┌──────────────────────────────────────┐ │ Unified STAC catalog (700K+ items) │ │ spanning 68 collections │ └──────────────────────────────────────┘ │ ┌────────────┬────────────────────┼────────────────────┬─────────────┐ ▼ ▼ ▼ ▼ ▼ NOAA ERDDAP IOOS RAs World Ocean DB Argo / GADR Sentinel / CoastWatch NDBC / CO-OPS Global Drifter OOI / IFREMER GOES-R / GOES-R MARACOOS, ... Program (GDP) EMODnet, ONC NEXRAD, ...

The server speaks theSTAC APIunder the hood.search_datasetscompiles your natural-language scope into a CQL2 filter;aggregateruns server-side bucket queries so the LLM never has to fetch raw items just to count them.

- More example walkthroughs (target: 30+)
- Reference integrations: LangChain, LlamaIndex, AutoGen, Mastra
- Per-source quickstart pages
- A "build your own MCP-powered ocean app" tutorial

Examples, prompt recipes, and integration guides are very welcome — seeCONTRIBUTING.md. Bug reports and source-coverage requests go inIssues.

Keywords: MCP server, Model Context Protocol, ocean data, atmospheric data, NOAA, NDBC, IOOS, World Ocean Database, Argo, GOES-R, NEXRAD, CoastWatch, ERDDAP, STAC, Claude, ChatGPT, Gemini, Grok, Cursor, AI agents, LLM tools, oceanography, climate data, marine data, satellite imagery, Sentinel.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Provides airfield aeronautical information for general aviation private pilots like runways, fuel prices, landing fees, weather, webcams and more

Swiss open data MCP server — transport, weather, geodata, companies, etc,. Zero API keys.

Weather data from the Mt Washington weather observatory in New Hampshire.

Weather forecasts, air quality, UV index, historical data, and alerts for AI agents. 5 tools, powered by Open-Meteo, free, unlimited, no API key.

Provides access to USDA SNOTEL weather and snow data.

Provides real-time and forecast WBGT (Heat Index) data for Tokyo from Japan's Ministry of the Environment.

Access earthquake data from the USGS Quakes API using natural language queries.

Real-time Amazon data API built for AI agents. 200M+ products, 1B+ reviews, live BSR, pricing, and competitor data as clean JSON. 10 agent skills for market research, competitor monitoring, pricing, listing audits, and more. 1,000 free credits.

Predict anything with Chronulus AI forecasting and prediction agents.

Get access to Kaggle's datasets, models, competitions, notebook and benchmarks.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.