noaa-marine-mcp-server

Get Water Level

noaa_marine_get_water_level

Read-onlyIdempotent

Observed water level, real-time or historical, for a CO-OPS water-level station, paired with tide predictions for the same period so the residual (observed − predicted) shows storm surge when positive and anomalous drawdown when negative. The interval selects the cadence: 6min (default) is the full curve, hourly and high_low cover months to a year of the same series at far fewer rows, and daily_mean serves Great Lakes stations only. Each interval carries its own CO-OPS range ceiling — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean — rejected locally before the call, and only 6min and hourly report a residual. Observations and predictions are fetched independently, so when the prediction series is empty, predictions_status says whether CO-OPS has none for this station and range or the prediction fetch failed — the observed series returns either way, and residual_summary is present only when both series are and the cadence supports the join. A sensor outage leaves slots with no reading; those slots are dropped and counted in gaps_dropped, so the observed series is continuous across the range only when that count is absent. A range whose rows fit the response budget is returned whole; a longer one returns the leading rows, with rows_matched, rows_returned, and next_offset reporting how much matched and which offset reaches the rest. Use noaa_marine_find_stations first to resolve a station name or location to a valid station ID.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`datum`	No	Datum the observed heights are referenced to. MLLW (default) is the US nautical chart datum, and MHHW (mean higher high water) is the flooding reference; MHW, MTL, MSL, and MLW are the other tidal planes. NAVD is NAVD88 and reads only where the station has an NAVD88 tie. STND is the station's own datum and is the plane that works where no tidal datum exists. IGLD and LWD apply at Great Lakes stations only, CRD at Columbia River stations only, and LWI and HWI are lunitidal intervals rather than heights. A datum this station does not carry comes back as datum_unavailable naming the ones it does.	MLLW
`limit`	No	Maximum observation rows to return on this page, for a caller that wants fewer than the response budget allows. Omit for the largest page that fits; a value larger than the budget allows does not widen the page.
`units`	No	Unit system: english = feet; metric = meters.	english
`offset`	No	Row offset into the matched observation series, for reading a range whose rows do not fit one response. 0 (default) starts at the first row; pass the next_offset from a previous call to continue. An offset past the last row returns an empty page rather than an error.
`end_date`	Yes	End date (inclusive), YYYYMMDD or YYYY-MM-DD, e.g. "20240601" or "2024-06-01".
`interval`	No	Observed cadence. 6min (default) is the full curve, 31 days per request, the only one carrying a quality flag. hourly is hourly heights, 365 days per request, and reports a residual against the hourly prediction series. high_low is the observed high and low waters with their H/HH/L/LL classification, 365 days per request, and reports no residual because observed extremes do not fall on predicted extreme times. daily_mean is the daily mean water level, 3655 days per request, published at Great Lakes stations only and with no paired prediction series; a coastal station comes back as great_lakes_only.	6min
`time_zone`	No	Time zone for returned timestamps. lst_ldt = local standard/daylight time (default); gmt = UTC; lst = local standard time year-round. Overridden to lst on interval="daily_mean", which CO-OPS serves in local standard time only and otherwise stamps every row a day off.	lst_ldt
`begin_date`	Yes	Start date, YYYYMMDD or YYYY-MM-DD, e.g. "20240601" or "2024-06-01".
`station_id`	Yes	CO-OPS water-level station ID (numeric, e.g. "9447130" for Seattle). Obtain from noaa_marine_find_stations with types=["water_level"].

Output Schema

TableJSON Schema

Name	Required	Description
`datum`	No	Tidal datum used — echoed for correct interpretation of water heights.
`error`	No	Present when the call failed. Absent on success.
`units`	No	Height units: "english" (feet) or "metric" (meters).
`notice`	No	Why the paired prediction series is missing or carries no residual, how many sample slots were dropped, and what this page of the observed series covers — whichever of those apply, composed into one string.
`interval`	No	Observed cadence echoed from the request — the CO-OPS product these rows came from, and what decides whether quality, type, and residual_summary are populated.
`truncated`	No	True when this page stops short of the last matched observation, so further rows remain at next_offset. False on the final page and on a page past the end. Absent when the whole matched series was returned.
`station_id`	No	Station ID echoed from the request — for chaining.
`next_offset`	No	Offset to pass as offset on the next call, or null when this page reaches the last matched observation. Absent when the whole matched series was returned.
`page_offset`	No	Observation row offset this page starts at, echoed from the request. Absent when the whole matched series was returned.
`predictions`	No	Paired tide predictions covering the same period as the returned observations, at the interval matching the observed cadence. Always empty on the daily_mean interval, which has no paired series. Otherwise empty when CO-OPS returned no predictions for this station and range, or when the prediction fetch failed — predictions_status says which.
`gaps_dropped`	No	Number of sample slots CO-OPS sent with no reading, dropped from observations. Present only when there was at least one. The observed series is therefore not continuous across the requested range: it has this many missing slots, and rows_matched counts only the slots that carry a value.
`observations`	No	Observed water level readings at the requested cadence — the whole matched series when it fits one response, otherwise the leading page starting at offset. rows_matched and next_offset report what a page left behind.
`rows_matched`	No	Observation rows the requested date range matched, before the page was cut. Absent when the whole matched series was returned.
`station_name`	No	Station name as returned by CO-OPS.
`rows_returned`	No	Observation rows this page carries — the length of observations. Absent when the whole matched series was returned.
`residual_summary`	No	Summary of observed-minus-predicted residuals in the requested units, computed across the whole matched series rather than the returned page — so a paged response still reports the largest surge in the range. Present only on the 6min and hourly intervals, and only when both series are available; the notice says why it is absent on high_low and daily_mean.
`predictions_status`	No	Present only when predictions is empty on an interval that has a paired series. "empty" means CO-OPS returned no prediction rows for this station and date range; "unavailable" means the prediction fetch failed, so no comparison series could be retrieved and the absence says nothing about the station. Absent when predictions were returned, and on the daily_mean interval, which has no paired series to report on.

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedSep 23, 2026

changedOutput schema / properties / error / properties / data / properties / reason / description

Previous value: -"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date. `date_range_exceeded`: Requested date range exceeds the selected interval's CO-OPS limit — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean. `no_data`: Station exists but no observed water-level data for the date range. `great_lakes_only`: interval=\"daily_mean\" was requested for a coastal station — CO-OPS publishes water-level daily means at Great Lakes stations only. `verified_data_lag`: The hourly, high_low, or daily_mean product returned no rows for the window, which CO-OPS has not published yet. `datum_unavailable`: The station does not carry the requested datum — a Great Lakes station has no tidal datum, and NAVD reads only where the station has an NAVD88 tie. `upstream_throttled`: CO-OPS answered the observed-series request with HTTP 403, which it returns for about two minutes while it throttles a burst of requests from one address. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date. `date_range_exceeded`: Requested date range exceeds the selected interval's CO-OPS limit — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean. `no_data`: Station exists but no observed water-level data for the date range — including an hourly, high_low, or daily_mean window ending before the prior month, which is not a verification lag. `great_lakes_only`: interval=\"daily_mean\" was requested for a coastal station — CO-OPS publishes water-level daily means at Great Lakes stations only. `verified_data_lag`: The hourly, high_low, or daily_mean product returned no rows for a window ending on or after the first day of the prior month, which CO-OPS may not have verified yet. `datum_unavailable`: The station does not carry the requested datum — a Great Lakes station has no tidal datum, and NAVD reads only where the station has an NAVD88 tie. `upstream_throttled`: CO-OPS answered the observed-series request with HTTP 403, which it returns for about two minutes while it throttles a burst of requests from one address. Other values are possible when a failure originates below the handler."

Changed8 schema fields changedSep 22, 2026

changedInput schema / properties / begin_date / description

Previous value: -"Start date in YYYYMMDD format, e.g. \"20240601\"."New value: +"Start date, YYYYMMDD or YYYY-MM-DD, e.g. \"20240601\" or \"2024-06-01\"."

changedInput schema / properties / begin_date / pattern

Previous value: -"^\\d{8}$"New value: +"^(\\d{8}|\\d{4}-\\d{2}-\\d{2})$"

changedInput schema / properties / end_date / description

Previous value: -"End date in YYYYMMDD format (inclusive), e.g. \"20240601\"."New value: +"End date (inclusive), YYYYMMDD or YYYY-MM-DD, e.g. \"20240601\" or \"2024-06-01\"."

changedInput schema / properties / end_date / pattern

Previous value: -"^\\d{8}$"New value: +"^(\\d{8}|\\d{4}-\\d{2}-\\d{2})$"

changedOutput schema / properties / error / properties / data / properties / reason / description

Previous value: -"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date `date_range_exceeded`: Requested date range exceeds the selected interval's CO-OPS limit — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean. `no_data`: Station exists but no observed water-level data for the date range. `great_lakes_only`: interval=\"daily_mean\" was requested for a coastal station — CO-OPS publishes water-level daily means at Great Lakes stations only. `verified_data_lag`: The hourly, high_low, or daily_mean product returned no rows for the window, which CO-OPS has not published yet. `datum_unavailable`: The station does not carry the requested datum — a Great Lakes station has no tidal datum, and NAVD reads only where the station has an NAVD88 tie. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date. `date_range_exceeded`: Requested date range exceeds the selected interval's CO-OPS limit — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean. `no_data`: Station exists but no observed water-level data for the date range. `great_lakes_only`: interval=\"daily_mean\" was requested for a coastal station — CO-OPS publishes water-level daily means at Great Lakes stations only. `verified_data_lag`: The hourly, high_low, or daily_mean product returned no rows for the window, which CO-OPS has not published yet. `datum_unavailable`: The station does not carry the requested datum — a Great Lakes station has no tidal datum, and NAVD reads only where the station has an NAVD88 tie. `upstream_throttled`: CO-OPS answered the observed-series request with HTTP 403, which it returns for about two minutes while it throttles a burst of requests from one address. Other values are possible when a failure originates below the handler."

changedOutput schema / properties / error / properties / data / properties / reason / examples

Previous value: -[
-  "station_not_found",
-  "invalid_date_range",
-  "date_range_exceeded",
-  "no_data",
-  "great_lakes_only",
-  "verified_data_lag",
-  "datum_unavailable"
-]New value: +[
+  "station_not_found",
+  "invalid_date_range",
+  "date_range_exceeded",
+  "no_data",
+  "great_lakes_only",
+  "verified_data_lag",
+  "datum_unavailable",
+  "upstream_throttled"
+]

changedOutput schema / properties / residual_summary / properties / max_drawdown / description

Previous value: -"Maximum negative residual magnitude in the requested units (feet for english, meters for metric) — anomalous drawdown indicator."New value: +"Maximum negative residual magnitude in the requested units (feet for english, meters for metric) — anomalous drawdown indicator. 0 when the observed level never fell below prediction in the range."

changedOutput schema / properties / residual_summary / properties / max_surge / description

Previous value: -"Maximum positive residual (observed − predicted) in the requested units (feet for english, meters for metric) — storm surge indicator."New value: +"Maximum positive residual (observed − predicted) in the requested units (feet for english, meters for metric) — storm surge indicator. 0 when the observed level never rose above prediction in the range."

Changed28 schema fields changedSep 17, 2026

changedInput schema / properties / datum / description

Previous value: -"Tidal datum reference plane. MLLW (default) is the US nautical chart datum. MSL = mean sea level; MHHW = mean higher high water (flooding reference)."New value: +"Datum the observed heights are referenced to. MLLW (default) is the US nautical chart datum, and MHHW (mean higher high water) is the flooding reference; MHW, MTL, MSL, and MLW are the other tidal planes. NAVD is NAVD88 and reads only where the station has an NAVD88 tie. STND is the station's own datum and is the plane that works where no tidal datum exists. IGLD and LWD apply at Great Lakes stations only, CRD at Columbia River stations only, and LWI and HWI are lunitidal intervals rather than heights. A datum this station does not carry comes back as datum_unavailable naming the ones it does."

changedInput schema / properties / datum / enum

Previous value: -[
-  "MLLW",
-  "MHHW",
-  "MSL",
-  "MTL",
-  "MHW",
-  "MLW",
-  "CD",
-  "STND"
-]New value: +[
+  "MLLW",
+  "MHHW",
+  "MHW",
+  "MTL",
+  "MSL",
+  "MLW",
+  "NAVD",
+  "STND",
+  "IGLD",
+  "LWD",
+  "CRD",
+  "LWI",
+  "HWI"
+]

addedInput schema / properties / interval

Added value: +{
+  "default": "6min",
+  "description": "Observed cadence. 6min (default) is the full curve, 31 days per request, the only one carrying a quality flag. hourly is hourly heights, 365 days per request, and reports a residual against the hourly prediction series. high_low is the observed high and low waters with their H/HH/L/LL classification, 365 days per request, and reports no residual because observed extremes do not fall on predicted extreme times. daily_mean is the daily mean water level, 3655 days per request, published at Great Lakes stations only and with no paired prediction series; a coastal station comes back as great_lakes_only.",
+  "enum": [
+    "6min",
+    "hourly",
+    "high_low",
+    "daily_mean"
+  ],
+  "type": "string"
+}

addedInput schema / properties / limit

Added value: +{
+  "description": "Maximum observation rows to return on this page, for a caller that wants fewer than the response budget allows. Omit for the largest page that fits; a value larger than the budget allows does not widen the page.",
+  "maximum": 9007199254740991,
+  "minimum": 1,
+  "type": "integer"
+}

addedInput schema / properties / offset

Added value: +{
+  "default": 0,
+  "description": "Row offset into the matched observation series, for reading a range whose rows do not fit one response. 0 (default) starts at the first row; pass the next_offset from a previous call to continue. An offset past the last row returns an empty page rather than an error.",
+  "maximum": 9007199254740991,
+  "minimum": 0,
+  "type": "integer"
+}

changedInput schema / properties / time_zone / description

Previous value: -"Time zone for returned timestamps. lst_ldt = local standard/daylight time (default); gmt = UTC; lst = local standard time year-round."New value: +"Time zone for returned timestamps. lst_ldt = local standard/daylight time (default); gmt = UTC; lst = local standard time year-round. Overridden to lst on interval=\"daily_mean\", which CO-OPS serves in local standard time only and otherwise stamps every row a day off."

changedOutput schema / anyOf

Previous value: -[
-  {
-    "not": {
-      "required": [
-        "error"
-      ]
-    },
-    "required": [
-      "station_id",
-      "station_name",
-      "datum",
-      "units",
-      "observations",
-      "predictions"
-    ]
-  },
-  {
-    "required": [
-      "error"
-    ]
-  }
-]New value: +[
+  {
+    "not": {
+      "required": [
+        "error"
+      ]
+    },
+    "required": [
+      "station_id",
+      "station_name",
+      "datum",
+      "units",
+      "interval",
+      "observations",
+      "predictions"
+    ]
+  },
+  {
+    "required": [
+      "error"
+    ]
+  }
+]

changedOutput schema / properties / error / properties / data / properties / reason / description

Previous value: -"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date `date_range_exceeded`: Requested date range exceeds the 31-day CO-OPS limit for 6-minute water level data. `no_data`: Station exists but no observed water-level data for the date range. Other values are possible when a failure originates below the handler."New value: +"Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date `date_range_exceeded`: Requested date range exceeds the selected interval's CO-OPS limit — 31 days for 6min, 365 for hourly and high_low, 3655 for daily_mean. `no_data`: Station exists but no observed water-level data for the date range. `great_lakes_only`: interval=\"daily_mean\" was requested for a coastal station — CO-OPS publishes water-level daily means at Great Lakes stations only. `verified_data_lag`: The hourly, high_low, or daily_mean product returned no rows for the window, which CO-OPS has not published yet. `datum_unavailable`: The station does not carry the requested datum — a Great Lakes station has no tidal datum, and NAVD reads only where the station has an NAVD88 tie. Other values are possible when a failure originates below the handler."

changedOutput schema / properties / error / properties / data / properties / reason / examples

Previous value: -[
-  "station_not_found",
-  "invalid_date_range",
-  "date_range_exceeded",
-  "no_data"
-]New value: +[
+  "station_not_found",
+  "invalid_date_range",
+  "date_range_exceeded",
+  "no_data",
+  "great_lakes_only",
+  "verified_data_lag",
+  "datum_unavailable"
+]

addedOutput schema / properties / gaps_dropped

Added value: +{
+  "description": "Number of sample slots CO-OPS sent with no reading, dropped from observations. Present only when there was at least one. The observed series is therefore not continuous across the requested range: it has this many missing slots, and rows_matched counts only the slots that carry a value.",
+  "maximum": 9007199254740991,
+  "minimum": -9007199254740991,
+  "type": "integer"
+}

addedOutput schema / properties / interval

Added value: +{
+  "description": "Observed cadence echoed from the request — the CO-OPS product these rows came from, and what decides whether quality, type, and residual_summary are populated.",
+  "enum": [
+    "6min",
+    "hourly",
+    "high_low",
+    "daily_mean"
+  ],
+  "type": "string"
+}

addedOutput schema / properties / next_offset

Added value: +{
+  "anyOf": [
+    {
+      "maximum": 9007199254740991,
+      "minimum": -9007199254740991,
+      "type": "integer"
+    },
+    {
+      "type": "null"
+    }
+  ],
+  "description": "Offset to pass as offset on the next call, or null when this page reaches the last matched observation. Absent when the whole matched series was returned."
+}

changedOutput schema / properties / notice / description

Previous value: -"Why the paired prediction series is missing, when it is missing."New value: +"Why the paired prediction series is missing or carries no residual, how many sample slots were dropped, and what this page of the observed series covers — whichever of those apply, composed into one string."

changedOutput schema / properties / observations / description

Previous value: -"6-minute observed water level readings."New value: +"Observed water level readings at the requested cadence — the whole matched series when it fits one response, otherwise the leading page starting at offset. rows_matched and next_offset report what a page left behind."

changedOutput schema / properties / observations / items / description

Previous value: -"A single 6-minute observed water level reading."New value: +"A single observed water level reading at the requested cadence."

changedOutput schema / properties / observations / items / properties / quality / description

Previous value: -"Quality flag: p = preliminary, v = verified."New value: +"Quality flag: p = preliminary, v = verified. Present on the 6min interval only — CO-OPS sends no flag with the hourly, high_low, or daily_mean products, which are verified data, so the field is absent rather than defaulted to preliminary."

changedOutput schema / properties / observations / items / properties / sigma / description

Previous value: -"Standard deviation of the water level sensor reading."New value: +"Standard deviation of the water level sensor reading. Present on the 6min and hourly intervals; CO-OPS sends none for high_low or daily_mean."

addedOutput schema / properties / observations / items / properties / type

Added value: +{
+  "description": "Observed extreme classification on the high_low interval: H = high water, HH = higher high water, L = low water, LL = lower low water. Absent on every other interval.",
+  "enum": [
+    "H",
+    "HH",
+    "L",
+    "LL"
+  ],
+  "type": "string"
+}

changedOutput schema / properties / observations / items / required

Previous value: -[
-  "time",
-  "value",
-  "quality"
-]New value: +[
+  "time",
+  "value"
+]

addedOutput schema / properties / page_offset

Added value: +{
+  "description": "Observation row offset this page starts at, echoed from the request. Absent when the whole matched series was returned.",
+  "maximum": 9007199254740991,
+  "minimum": -9007199254740991,
+  "type": "integer"
+}

changedOutput schema / properties / predictions / description

Previous value: -"Paired 6-minute tide predictions for the same period. Empty when CO-OPS returned no predictions for this station and range, or when the prediction fetch failed — predictions_status says which."New value: +"Paired tide predictions covering the same period as the returned observations, at the interval matching the observed cadence. Always empty on the daily_mean interval, which has no paired series. Otherwise empty when CO-OPS returned no predictions for this station and range, or when the prediction fetch failed — predictions_status says which."

changedOutput schema / properties / predictions / items / description

Previous value: -"A single 6-minute tide prediction."New value: +"A single tide prediction."

changedOutput schema / properties / predictions / items / properties / time / description

Previous value: -"Prediction datetime matching the observation time step."New value: +"Prediction datetime in the requested time zone."

changedOutput schema / properties / predictions_status / description

Previous value: -"Present only when predictions is empty. \"empty\" means CO-OPS returned no prediction rows for this station and date range; \"unavailable\" means the prediction fetch failed, so no comparison series could be retrieved and the absence says nothing about the station. Absent when predictions were returned."New value: +"Present only when predictions is empty on an interval that has a paired series. \"empty\" means CO-OPS returned no prediction rows for this station and date range; \"unavailable\" means the prediction fetch failed, so no comparison series could be retrieved and the absence says nothing about the station. Absent when predictions were returned, and on the daily_mean interval, which has no paired series to report on."

changedOutput schema / properties / residual_summary / description

Previous value: -"Summary of observed-minus-predicted residuals in the requested units. Only present when both observations and predictions are available."New value: +"Summary of observed-minus-predicted residuals in the requested units, computed across the whole matched series rather than the returned page — so a paged response still reports the largest surge in the range. Present only on the 6min and hourly intervals, and only when both series are available; the notice says why it is absent on high_low and daily_mean."

addedOutput schema / properties / rows_matched

Added value: +{
+  "description": "Observation rows the requested date range matched, before the page was cut. Absent when the whole matched series was returned.",
+  "maximum": 9007199254740991,
+  "minimum": -9007199254740991,
+  "type": "integer"
+}

addedOutput schema / properties / rows_returned

Added value: +{
+  "description": "Observation rows this page carries — the length of observations. Absent when the whole matched series was returned.",
+  "maximum": 9007199254740991,
+  "minimum": -9007199254740991,
+  "type": "integer"
+}

addedOutput schema / properties / truncated

Added value: +{
+  "description": "True when this page stops short of the last matched observation, so further rows remain at next_offset. False on the final page and on a page past the end. Absent when the whole matched series was returned.",
+  "type": "boolean"
+}

Changed3 schema fields changedSep 17, 2026

addedOutput schema / properties / notice

Added value: +{
+  "description": "Why the paired prediction series is missing, when it is missing.",
+  "type": "string"
+}

changedOutput schema / properties / predictions / description

Previous value: -"Paired 6-minute tide predictions for the same period. May be empty if CO-OPS predictions are unavailable for this station."New value: +"Paired 6-minute tide predictions for the same period. Empty when CO-OPS returned no predictions for this station and range, or when the prediction fetch failed — predictions_status says which."

addedOutput schema / properties / predictions_status

Added value: +{
+  "description": "Present only when predictions is empty. \"empty\" means CO-OPS returned no prediction rows for this station and date range; \"unavailable\" means the prediction fetch failed, so no comparison series could be retrieved and the absence says nothing about the station. Absent when predictions were returned.",
+  "enum": [
+    "empty",
+    "unavailable"
+  ],
+  "type": "string"
+}

Changed6 schema fields changedAug 25, 2026

changedInput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

addedInput schema / additionalProperties
```
Added value: +false
```

changedOutput schema / $schema

Previous value: -"http://json-schema.org/draft-07/schema#"New value: +"https://json-schema.org/draft/2020-12/schema"

addedOutput schema / anyOf

Added value: +[
+  {
+    "not": {
+      "required": [
+        "error"
+      ]
+    },
+    "required": [
+      "station_id",
+      "station_name",
+      "datum",
+      "units",
+      "observations",
+      "predictions"
+    ]
+  },
+  {
+    "required": [
+      "error"
+    ]
+  }
+]

addedOutput schema / properties / error

Added value: +{
+  "additionalProperties": {},
+  "description": "Present when the call failed. Absent on success.",
+  "properties": {
+    "code": {
+      "description": "JSON-RPC error code for this failure.",
+      "maximum": 9007199254740991,
+      "minimum": -9007199254740991,
+      "type": "integer"
+    },
+    "data": {
+      "additionalProperties": {},
+      "properties": {
+        "reason": {
+          "description": "Machine-readable failure mode. Declared by this tool: `station_not_found`: CO-OPS returned an error for the station ID. `invalid_date_range`: begin_date/end_date is not a real calendar date or begin_date is after end_date `date_range_exceeded`: Requested date range exceeds the 31-day CO-OPS limit for 6-minute water level data. `no_data`: Station exists but no observed water-level data for the date range. Other values are possible when a failure originates below the handler.",
+          "examples": [
+            "station_not_found",
+            "invalid_date_range",
+            "date_range_exceeded",
+            "no_data"
+          ],
+          "type": "string"
+        },
+        "recovery": {
+          "additionalProperties": {},
+          "description": "Actionable next step for the caller.",
+          "properties": {
+            "hint": {
+              "type": "string"
+            }
+          },
+          "required": [
+            "hint"
+          ],
+          "type": "object"
+        },
+        "retryable": {
+          "description": "Whether retrying may succeed.",
+          "type": "boolean"
+        }
+      },
+      "type": "object"
+    },
+    "message": {
+      "description": "Human-readable description of what went wrong.",
+      "type": "string"
+    }
+  },
+  "required": [
+    "code",
+    "message"
+  ],
+  "type": "object"
+}

removedOutput schema / required

Removed value: -[
-  "station_id",
-  "station_name",
-  "datum",
-  "units",
-  "observations",
-  "predictions"
-]

First observedJun 7, 2026

TDQS

A4.6/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint, idempotentHint, and openWorldHint, so the safety profile is covered; the description then adds substantial behavior beyond that: independent fetching of observations and predictions, predictions_status semantics on empty series, gaps_dropped behavior for sensor outages, local rejection of range ceilings, datum_unavailable and great_lakes_only failure modes, and offset-based pagination with rows_matched/rows_returned/next_offset. No contradiction with the annotations — this is a pure read.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The core purpose is front-loaded in sentence one, and every subsequent sentence carries distinct operational information about cadence, limits, gaps, or paging — no filler. It is a long block for the agent to parse, and the interval sentence partially overlaps with the schema's interval enum description, but for a 9-parameter tool with subtle boundary semantics the length is mostly earned.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return-value details are covered elsewhere; the description supplies everything else an agent needs to call correctly: the station-resolution prerequisite, per-interval range ceilings with local validation, residual availability rules, gap handling, pagination continuation, and the edge-case responses (datum_unavailable, great_lakes_only, predictions_status). Nothing material is missing for correct invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and the individual parameter descriptions are unusually rich (datum planes explained, interval residuals per cadence, time-zone override on daily_mean). The description still adds meaning beyond the schema by tying parameters together: the range-ceiling ceiling per interval, which intervals support the residual join, and how limit/offset connect to the paginated response fields.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The first sentence states a specific verb and resource: observed water level for a CO-OPS station paired with tide predictions, with residual defined as observed minus predicted. The storm-surge/drawdown framing makes the tool's reason for existing concrete, and it is clearly distinguishable from siblings like noaa_marine_get_tide_predictions (predictions only) and noaa_marine_get_monthly_means.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly tells the agent to run noaa_marine_find_stations first to resolve a station ID, and gives rich selection guidance per interval (daily_mean for Great Lakes only, 6min for the full curve, hourly/high_low for long ranges). However, it never explicitly states when-not-to-use this tool versus siblings such as noaa_marine_get_tide_predictions for prediction-only needs — that contrast is implied rather than named.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All noaa-marine-mcp-server tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Get Water Level

Input Schema

Output Schema

Schema Changelog

TDQS

Glama MCP Gateway

Resources