mcnoaa-tides

MCP server for NOAA CO-OPS Tides and Currents. Exposes tide predictions, observed water levels, and meteorological data from ~301 U.S. coastal stations as tools, resources, and prompts via FastMCP 3.0.

Built for marine planning — fishing trips, boating, crabbing, safety checks.

Install

# Run directly (no install needed)
uvx mcnoaa-tides

# Or add to Claude Code
claude mcp add mcnoaa-tides -- uvx mcnoaa-tides

# With visualization support (charts)
uv pip install mcnoaa-tides[viz]

# Local development
uv run mcnoaa-tides

Tools

search_stations — Find stations by name, state, or type

search_stations(state="WA")

Returns up to 50 matching stations:

[
  {"id": "9447130", "name": "Seattle", "state": "WA", "lat": 47.6026, "lng": -122.3393, "tidal": true},
  {"id": "9446484", "name": "Tacoma", "state": "WA", "lat": 47.2671, "lng": -122.4132, "tidal": true},
  {"id": "9444900", "name": "Port Townsend", "state": "WA", "lat": 48.1129, "lng": -122.7595, "tidal": true}
]

Also supports query (name search) and is_tidal (filter tidal vs non-tidal).

find_nearest_stations — Proximity search by coordinates

find_nearest_stations(latitude=47.6, longitude=-122.34, limit=3)

Distances in nautical miles:

[
  {"id": "9447130", "name": "Seattle", "state": "WA", "lat": 47.6026, "lng": -122.3393, "distance_nm": 0.2},
  {"id": "9446484", "name": "Tacoma", "state": "WA", "lat": 47.2671, "lng": -122.4132, "distance_nm": 20.1},
  {"id": "9445958", "name": "Bremerton", "state": "WA", "lat": 47.5615, "lng": -122.6225, "distance_nm": 14.7}
]

get_station_info — Expanded station metadata

get_station_info(station_id="9447130")

Returns sensors, datums, products, and station details. Seattle has been operating since 1899.

Full response (194 lines)

See examples/station-detail-seattle.json

get_tide_predictions — High/low tide times

get_tide_predictions(station_id="9447130", hours=24)

Defaults to hilo interval (high/low times only) — the most useful format for planning:

{
  "predictions": [
    {"t": "2026-02-21 00:49", "v": "2.658", "type": "L"},
    {"t": "2026-02-21 07:08", "v": "12.261", "type": "H"},
    {"t": "2026-02-21 13:43", "v": "1.167", "type": "L"},
    {"t": "2026-02-21 19:54", "v": "9.857", "type": "H"}
  ]
}

type: H = high tide, L = low tide. Values in feet above MLLW.

Other intervals: "h" (hourly), "6" (6-minute). Datum options: MLLW, MSL, NAVD, STND.

get_observed_water_levels — Actual readings

get_observed_water_levels(station_id="9447130", hours=3)

6-minute interval observations. Compare with predictions to see how reality diverges:

{
  "metadata": {"id": "9447130", "name": "Seattle", "lat": "47.6026", "lon": "-122.3393"},
  "data": [
    {"t": "2026-02-21 17:24", "v": "7.41", "s": "0.059", "f": "0,0,0,0", "q": "p"},
    {"t": "2026-02-21 17:30", "v": "7.62", "s": "0.072", "f": "0,0,0,0", "q": "p"},
    {"t": "2026-02-21 17:36", "v": "7.803", "s": "0.069", "f": "0,0,0,0", "q": "p"}
  ]
}

q: "p" = preliminary (real-time), "v" = verified (post-processed). s = standard deviation.

Full 3-hour response (213 lines)

See examples/water-levels-seattle.json

get_meteorological_data — Weather observations

One tool, 8 products. Select via product parameter:

Product	Fields	Units
air_temperature	v	deg F
water_temperature	v	deg F
wind	s (speed), d (dir deg), dr (compass), g (gust)	knots
air_pressure	v	millibars
conductivity	v	mS/cm
visibility	v	nautical miles
humidity	v	percent
salinity	v	PSU

get_meteorological_data(station_id="8454000", product="wind", hours=1)

{
  "metadata": {"id": "8454000", "name": "Providence", "lat": "41.8072", "lon": "-71.4007"},
  "data": [
    {"t": "2026-02-21 22:24", "s": "2.72", "d": "109.0", "dr": "ESE", "g": "3.3", "f": "0,0"},
    {"t": "2026-02-21 22:30", "s": "3.3", "d": "103.0", "dr": "ESE", "g": "3.89", "f": "0,0"},
    {"t": "2026-02-21 22:36", "s": "2.72", "d": "108.0", "dr": "ESE", "g": "3.5", "f": "0,0"}
  ]
}

Not all stations support all products — use get_station_info to check available sensors.

marine_conditions_snapshot — Everything at once

marine_conditions_snapshot(station_id="9447130")

Fires 6 parallel API calls (predictions, water level, water temp, air temp, wind, pressure) and returns a combined snapshot. Products that aren't available at a station appear under unavailable instead of failing the whole request:

{
  "station_id": "9447130",
  "fetched_utc": "2026-02-22T04:15:38.291Z",
  "predictions": {"predictions": [{"t": "...", "v": "12.261", "type": "H"}, "..."]},
  "water_level": {"data": [{"t": "...", "v": "10.22", "s": "0.053", "f": "0,0,0,0", "q": "p"}, "..."]},
  "air_pressure": {"data": [{"t": "...", "v": "1012.4", "f": "0,0,0"}, "..."]},
  "unavailable": {
    "water_temperature": "ValueError: No data was found...",
    "air_temperature": "ValueError: No data was found...",
    "wind": "ValueError: No data was found..."
  }
}

Best starting point for trip planning or safety assessments.

visualize_tides — Tide prediction chart

visualize_tides(station_id="9447130", hours=48, format="png")

Generates a tide prediction chart with the water level curve, high/low markers, and optional observed water level overlay. PNG returns an inline image via MCP; HTML saves an interactive Plotly chart to artifacts/charts/.

Parameters:

• hours — forecast window (default 48)
• include_observed — overlay actual readings (default true)
• format — "png" (inline image) or "html" (interactive file)

Requires mcnoaa-tides[viz] — install with uv pip install mcnoaa-tides[viz].

visualize_conditions — Multi-panel conditions dashboard

visualize_conditions(station_id="9447130", hours=24, format="png")

Generates a multi-panel dashboard with up to 4 panels: tide predictions + observed overlay, wind speed/gust, air/water temperature, and barometric pressure with trend indicator. Products unavailable at a station are omitted.

Parameters:

• hours — data window (default 24)
• format — "png" (inline image) or "html" (interactive file)

Requires mcnoaa-tides[viz].

Resources

URI	Description
noaa://stations	Full station catalog (~301 stations)
noaa://stations/{station_id}	Expanded metadata for one station
noaa://stations/{station_id}/nearby	Stations within 50 nm

Prompts

plan_fishing_trip

Guides the LLM through station discovery, tide analysis, and weather assessment to recommend optimal fishing windows.

plan_fishing_trip(location="Narragansett Bay", target_species="striped bass", date="20260615")

marine_safety_check

GO / CAUTION / NO-GO assessment based on wind, visibility, water temperature, and pressure trends.

marine_safety_check(station_id="9447130")

Response field reference

Field	Meaning
t	Timestamp (local station time)
v	Value (water level ft, temp F, pressure mb, etc.)
type	H (high) or L (low) — predictions only
s	Speed (wind, knots) or sigma (water level, std deviation)
d	Wind direction in degrees true
dr	Wind compass direction (N, NE, SW, etc.)
g	Wind gust speed (knots)
f	Data quality flags (comma-separated)
q	QA level: p = preliminary, v = verified

Development

git clone <repo-url> && cd mcnoaa-tides
uv sync --dev

# Run tests (mock client, no network)
uv run pytest tests/ -v

# Lint
uv run ruff check src/

# Start server locally
uv run mcnoaa-tides

# Headless test with Claude
claude -p "Search for tide stations in Rhode Island" \
  --mcp-config .mcp.json \
  --allowedTools "mcp__mcnoaa-tides__*"

Data source

All data from NOAA Center for Operational Oceanographic Products and Services (CO-OPS). No API key required. Station IDs are 7-digit numbers (e.g. 9447130 for Seattle, 8454000 for Providence).

Two separate APIs are used:

• Data API — observations and predictions (api.tidesandcurrents.noaa.gov/api/prod/datagetter)
• Metadata API — station info, sensors, datums (api.tidesandcurrents.noaa.gov/mdapi/prod/webapi)

License

MIT