Climate risk analytics - Cotality developer portal

Data capabilities:

Dataset	Key metrics	Use cases
Age of roof	Estimated roof age (years), modeled construction year, confidence rank (statistical estimate — not a county record)	Insurance underwriting, hazard policy pricing, roof replacement planning
Climate risk	Per-peril modelled loss and risk scores (perils include flood, wildfire, wind, and earthquake) at 50-, 100-, 200-, and 500-year return periods, under a current baseline and forward climate scenarios	Multi-peril exposure modeling, portfolio risk concentration, compliance reporting, reinsurance pricing

Data scope:

Attribute	Value
Input	One or more CLIP identifiers — a CLIP is Cotality’s unique, persistent numeric identifier for a US parcel. Use CLIP tools to resolve an address to a CLIP first.
Output	MCP tool envelope — `data` (property array with age of roof and/or climate risk payload) and `count`
Coverage	US residential properties
Update frequency	Age of roof: refreshed periodically from aerial and satellite imagery. Climate risk: refreshed annually. Neither runs to a published calendar — read `modelRunDate` and `modelVersion` on the response to see when the data you received was produced.
Exclusions	Commercial real estate is not covered. Age of Roof is a statistical model estimate — always check `attributes.yearBuiltModeledIndicator` on the roof age item; a value of `1` means the year is modeled, not recorded.
Combine with	Property characteristics for structural data; Analytics for market context

Recommended tool call sequence

Start from a CLIP. If you only have an address, resolve it first — these tools do not accept addresses.

clip-find_property_by_address   → resolve the address, capture data[0].clip
at-get_property_analytics       → roof age + climate risk in one call
   (or at-get_property_roof_age / at-get_property_climate_risk for one dataset)
pd-get_property_characteristics → structural detail, if you need it

Prefer at-get_property_analytics when you want both datasets — it is one request against your quota instead of two. Use the single-dataset tools when your token holds only one of the two scopes.

`at-get_property_analytics`

Required scope: risk:analytics. Retrieve comprehensive property analytics data — both age of roof and climate risk — for a list of CLIPs in a single call. When to use:

You need both roof condition and climate risk data for one or more properties
Performing full property risk assessments that combine physical condition with climate exposure
Running portfolio-level analyses where you want all available analytics in one request

Examples:

“Get full analytics for CLIPs 123456789 and 987654321”
“Retrieve roof age and climate risk for CLIP 555555555”

Input schema

Parameter	Type	Required	Description
`clips`	`array[string]`	Yes	A list of numeric CLIP identifier strings. Minimum 1, maximum 50 per request. Each item must match `^\d+$`.

Each CLIP must be a non-empty string consisting only of numeric characters (e.g. ["123456789", "987654321"])

Example request

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "at-get_property_analytics",
    "arguments": {
      "clips": ["123456789"]
    }
  }
}

Output schema

Each item in data carries clip, a roofAge array of RoofAgeItem, and a climateRisk array of ClimateRiskItem. Both arrays are flat — there is no propertyAnalytics, ageOfRoof, or climateRiskAnalytics wrapper. See Response envelope and Response models.

`at-get_property_roof_age`

Required scope: risk:age-of-roof. Retrieve age of roof data only — roof age, construction year, confidence rank, and property address with geocode — for a list of CLIPs. When to use:

You only need roof condition data without climate risk scores
Monitoring roof age across a portfolio for replacement planning
Validating roof construction year during insurance underwriting

Examples:

“Get roof age for CLIP 123456789”
“Check roof construction year and confidence for CLIPs 111111111 and 222222222”

Input schema

Parameter	Type	Required	Description
`clips`	`array[string]`	Yes	A list of numeric CLIP identifier strings. Minimum 1, maximum 50 per request. Each item must match `^\d+$`.

Each CLIP must be a non-empty string consisting only of numeric characters (e.g. ["123456789", "987654321"])

Example request

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "at-get_property_roof_age",
    "arguments": {
      "clips": ["123456789"]
    }
  }
}

Output schema

Each item in data carries clip and a roofAge array of RoofAgeItem. There is no climateRisk array and no propertyAnalytics or ageOfRoof wrapper. See Response envelope and Roof age models.

`at-get_property_climate_risk`

Required scope: risk:climate-analytics. Retrieve climate risk data only — multi-peril loss scores, climate scenario comparisons, and projections by year — for a list of CLIPs. The server returns one flat perils[] row for every peril, climate scenario, and projection year combination. The same peril therefore appears many times in the array. The set of perils returned varies by property and by model version. Read perils[].description and perils[].perilId from the response rather than assuming a fixed list.

Climate scenario — a standardised assumption about future greenhouse gas concentrations, used to project risk forward. Cotality returns Representative Concentration Pathways (RCP 2.6 “favorable”, RCP 4.5 “moderate”, RCP 8.5 “severe”) and Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, SSP5-8.5), plus a BASE baseline representing current conditions. See PerilItem for the exact codes.

When to use:

You only need climate risk exposure without roof condition data
Evaluating climate resilience for one or more properties
Comparing future risk projections across climate scenarios and projection years

Examples:

“Get climate risk scores for CLIP 123456789”
“Compare flood and wildfire risk for CLIPs 111111111 and 222222222”

Input schema

Parameter	Type	Required	Description
`clips`	`array[string]`	Yes	A list of numeric CLIP identifier strings. Minimum 1, maximum 50 per request. Each item must match `^\d+$`.

Each CLIP must be a non-empty string consisting only of numeric characters (e.g. ["123456789", "987654321"])

Example request

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "at-get_property_climate_risk",
    "arguments": {
      "clips": ["123456789"]
    }
  }
}

Output schema

Each item in data carries clip and a climateRisk array of ClimateRiskItem. There is no roofAge array and no propertyAnalytics or climateRiskAnalytics wrapper. See Response envelope and Climate risk models.

Response envelope

All three tools return an identical envelope:

Field	Type	Description
`data`	`array`	Array of property records (item type varies by tool — see below)
`count`	`integer`	Number of records in `data`

Both fields are always present in a successful response. There is no success or messages field — failure is signalled by an MCP error carrying a ToolError payload. See Response envelope and MCP tool errors.

A CLIP that matches no property is not an error. You receive a successful response with count of 0 and an empty data array. When you submit several CLIPs, only the ones that matched are returned — compare count against the number you sent.

Response models

Each tool returns one flat item per CLIP. The upstream propertyAnalytics, ageOfRoof, and climateRiskAnalytics wrappers are removed before the data reaches you, and address, geocode, and per-scenario values are promoted or flattened. All fields are optional and are omitted when no value is available.

Tool	Item fields
`at-get_property_analytics`	`clip`, `roofAge`, `climateRisk`
`at-get_property_roof_age`	`clip`, `roofAge`
`at-get_property_climate_risk`	`clip`, `climateRisk`

Field	Type	Description
`clip`	`string`	Cotality CLIP identifier for the property
`roofAge`	`array[RoofAgeItem]`	Roof age records (see RoofAgeItem)
`climateRisk`	`array[ClimateRiskItem]`	Climate risk records (see ClimateRiskItem)

Do not look for propertyAnalytics, ageOfRoof, climateRiskAnalytics, craStructureGeocode, rcpScenarios, or timeHorizons in the response. Those names belong to the upstream data model, not to the MCP tool output.

Roof age models

RoofAgeItem

Each RoofAgeItem holds one roof estimate, with the address and geocode fields promoted to the top of the item. All fields are optional and are omitted when no value is available.

Field	Type	Description
`address1`	`string`	Primary address line
`city`	`string`	City name
`state`	`string`	State abbreviation
`zipCode`	`string`	5-digit ZIP code
`streetName`	`string`	Street name
`latitude`	`number`	Latitude coordinate
`longitude`	`number`	Longitude coordinate
`roofConstructedAge`	`integer`	Estimated age of the roof in years
`roofConstructedConfidenceRank`	`string`	Confidence in the roof construction estimate: `High`, `High_modeled`, `Medium_modeled`, or `Low_modeled`
`roofConstructedYear`	`integer`	Estimated year the roof was constructed or last replaced
`yearBuilt`	`integer`	Year the building was originally constructed
`attributes`	`RoofAgeAttributes`	Lower-frequency fields (see RoofAgeAttributes)

RoofAgeItem does not repeat clip. Read the CLIP from the parent item in data.

RoofAgeAttributes

The attributes object groups the lower-frequency roof fields. It is omitted entirely when none of its fields has a value.

Field	Type	Description
`medianRoofConstructedAge`	`integer`	Median roof construction age across properties at this location
`medianRoofConstructedYear`	`integer`	Median roof construction year across properties at this location
`totalNumberOfProperties`	`integer`	Total number of properties evaluated at this location
`yearBuiltModeledIndicator`	`integer`	Indicator for whether the year built value is modeled (1) or recorded (0)

Understanding `roofConstructedConfidenceRank`

Each RoofAgeItem returns a roofConstructedConfidenceRank describing the model’s confidence in the estimated construction year — not in the age itself. Values are case-sensitive and must be compared exactly as returned.

Value	Meaning
`High`	Very high confidence
`High_modeled`	High confidence
`Medium_modeled`	Moderate confidence
`Low_modeled`	Low confidence

A Low_modeled estimate is not equivalent to missing data; it means the model produced an estimate but from weaker underlying signals. The field is omitted entirely when no value is available.

Match these values exactly, including capitalization and the _modeled suffix. Uppercase forms such as HIGH, MEDIUM, or LOW are never returned, and a comparison against them silently fails.

Climate risk models

ClimateRiskItem

Each ClimateRiskItem holds one climate risk model run, with the structure geocode promoted to the top of the item. All fields are optional and are omitted when no value is available.

Field	Type	Description
`latitude`	`number`	Latitude of the geocode used for the analysis
`longitude`	`number`	Longitude of the geocode used for the analysis
`craRecordIdentification`	`string`	Unique identifier for this climate risk analytics record
`modelRunDate`	`string`	Date the climate risk model was executed
`modelVersion`	`string`	Version of the climate risk model used
`riskTypeScoreGroup`	`string`	Property class the risk score is benchmarked against — `SFR` (single family), `MFR` (multi family), `COM` (commercial), `IND` (industrial), `MUN` (municipal), or `AGR` (agricultural)
`perils`	`array[PerilItem]`	Flattened per-peril risk rows (see PerilItem)

PerilItem

perils is a fully flattened list. The server expands every peril across every climate scenario and every projection year, and emits one row per combination. The same peril appears many times — once for each climateScenarioCode and timeHorizon pair.

Do not assume one row per peril. Group or filter perils on perilId + climateScenarioCode

timeHorizon before you compare or aggregate. Reading perils[0] gives you one arbitrary scenario-year combination, not the peril’s overall risk.

All fields are optional and are omitted when no value is available.

Field	Type	Description
`description`	`string`	Human-readable name of the peril (e.g. “Flood”, “Wildfire”)
`perilId`	`string`	Machine-readable peril identifier (e.g. `"EQSH"`, `"FIXX"`)
`climateScenarioCode`	`string`	Scenario identifier. Delivered without a decimal point — for example `"BASE"`, `"RCP26"`, `"RCP45"`, `"RCP85"`, `"SSP1_26"`, `"SSP2_45"`, `"SSP5_85"`.
`averageAnnualLossRatio`	`number`	Long-term average annual loss as a ratio of replacement cost
`averageAnnualLossRiskScore`	`integer`	`0`–`100`, ranking this property against all properties nationwide in the same `riskTypeScoreGroup`. `0` means the peril is not modelled at this location — it does not mean zero risk. Where the peril is modelled, `1` is the lowest risk and `100` the highest.
`timeHorizon`	`integer`	Projection year in `YYYY` format (e.g. `2030`, `2040`, `2050`). A value of `0` means current conditions, not the year zero.
`attributes`	`PerilAttributes`	Return-period loss estimates (see PerilAttributes)

The peril list varies by property and model version. Read description and perilId rather than coding against a fixed set.

Climate scenario codes

Code	Meaning
`BASE`	Current conditions — the baseline, not a projection
`RCP26` / `SSP1_26`	Favorable climate response
`RCP45` / `SSP2_45`	Moderate climate response
`RCP85` / `SSP5_85`	Severe climate response

Do not parse climateScenarioCode as a number. "RCP45" means RCP 4.5, not 45. Match the code as a string against the table above.

PerilAttributes

The return-period loss estimates live inside attributes on each peril row — not directly on the peril. The object is omitted entirely when none of its fields has a value. Three concepts run through this object:

Term	Meaning
Return period	How rare the loss is. A 100-year return period is the loss level with a 1-in-100 (1%) chance of being exceeded in any given year.
AEP vs OEP	Annual Exceedance Probability is based on all losses in a year combined. Occurrence Exceedance Probability is based on the largest single loss that year. AEP is the higher number for perils that can strike repeatedly.
TVaR	Tail Value-at-Risk. Instead of the loss at the threshold, it is the average of all losses above it — the answer to “if it is worse than the 1-in-100-year event, how bad on average?”

The 200-year fields are populated for AR6 SSP scenarios only. Expect them to be absent under the RCP scenarios.

Field	Type	Description
`annualExceedanceProbability50YearLoss`	`number`	AEP loss estimate for a 50-year return period
`annualExceedanceProbability100YearLoss`	`number`	AEP loss estimate for a 100-year return period
`annualExceedanceProbability200YearLoss`	`number`	AEP loss estimate for a 200-year return period
`annualExceedanceProbability500YearLoss`	`number`	AEP loss estimate for a 500-year return period
`annualExceedanceProbability50YearLossTvar`	`number`	AEP Tail Value-at-Risk for a 50-year return period
`annualExceedanceProbability100YearLossTvar`	`number`	AEP Tail Value-at-Risk for a 100-year return period
`annualExceedanceProbability200YearLossTvar`	`number`	AEP Tail Value-at-Risk for a 200-year return period
`annualExceedanceProbability500YearLossTvar`	`number`	AEP Tail Value-at-Risk for a 500-year return period
`occurrenceExceedenceProbability50YearLoss`	`number`	OEP loss estimate for a 50-year return period
`occurrenceExceedenceProbability100YearLoss`	`number`	OEP loss estimate for a 100-year return period
`occurrenceExceedenceProbability200YearLoss`	`number`	OEP loss estimate for a 200-year return period
`occurrenceExceedenceProbability500YearLoss`	`number`	OEP loss estimate for a 500-year return period
`occurrenceExceedenceProbability50YearLossTvar`	`number`	OEP Tail Value-at-Risk for a 50-year return period
`occurrenceExceedenceProbability100YearLossTvar`	`number`	OEP Tail Value-at-Risk for a 100-year return period
`occurrenceExceedenceProbability200YearLossTvar`	`number`	OEP Tail Value-at-Risk for a 200-year return period
`occurrenceExceedenceProbability500YearLossTvar`	`number`	OEP Tail Value-at-Risk for a 500-year return period

The AEP fields spell it “Exceedance” and the OEP fields spell it “Exceedence”. That inconsistency comes from the upstream data model and is reproduced exactly in the tool output — it is not a typo in this page. Match the field names character for character.

Sample response — full analytics

at-get_property_analytics returns both arrays. This example shows one peril (FLOOD) expanded across two scenarios, and a second peril (WILDFIRE) under one scenario — four flat rows in perils for two perils.

{
  "count": 1,
  "data": [
    {
      "clip": "123456789",
      "roofAge": [
        {
          "address1": "123 MAIN ST",
          "city": "LOS ANGELES",
          "state": "CA",
          "zipCode": "90001",
          "streetName": "MAIN",
          "latitude": 34.0522,
          "longitude": -118.2437,
          "roofConstructedAge": 16,
          "roofConstructedConfidenceRank": "High",
          "roofConstructedYear": 2010,
          "yearBuilt": 1985,
          "attributes": {
            "medianRoofConstructedAge": 18,
            "medianRoofConstructedYear": 2008,
            "totalNumberOfProperties": 1,
            "yearBuiltModeledIndicator": 0
          }
        }
      ],
      "climateRisk": [
        {
          "latitude": 34.0522,
          "longitude": -118.2437,
          "craRecordIdentification": "CRA-123456789",
          "modelRunDate": "2025-01-15",
          "modelVersion": "3.0",
          "riskTypeScoreGroup": "SFR",
          "perils": [
            {
              "description": "Flood",
              "perilId": "FLOOD",
              "climateScenarioCode": "BASE",
              "averageAnnualLossRatio": 0.00041,
              "averageAnnualLossRiskScore": 11,
              "timeHorizon": 0,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.0021,
                "annualExceedanceProbability100YearLoss": 0.0047,
                "annualExceedanceProbability500YearLoss": 0.0131,
                "annualExceedanceProbability50YearLossTvar": 0.0035,
                "annualExceedanceProbability100YearLossTvar": 0.0067,
                "annualExceedanceProbability500YearLossTvar": 0.0176,
                "occurrenceExceedenceProbability50YearLoss": 0.0017,
                "occurrenceExceedenceProbability100YearLoss": 0.0041,
                "occurrenceExceedenceProbability500YearLoss": 0.0115,
                "occurrenceExceedenceProbability50YearLossTvar": 0.0029,
                "occurrenceExceedenceProbability100YearLossTvar": 0.0058,
                "occurrenceExceedenceProbability500YearLossTvar": 0.0157
              }
            },
            {
              "description": "Flood",
              "perilId": "FLOOD",
              "climateScenarioCode": "RCP45",
              "averageAnnualLossRatio": 0.00045,
              "averageAnnualLossRiskScore": 12,
              "timeHorizon": 2030,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.0023,
                "annualExceedanceProbability100YearLoss": 0.0051,
                "annualExceedanceProbability500YearLoss": 0.0142,
                "annualExceedanceProbability50YearLossTvar": 0.0038,
                "annualExceedanceProbability100YearLossTvar": 0.0072,
                "annualExceedanceProbability500YearLossTvar": 0.019,
                "occurrenceExceedenceProbability50YearLoss": 0.0019,
                "occurrenceExceedenceProbability100YearLoss": 0.0045,
                "occurrenceExceedenceProbability500YearLoss": 0.0125,
                "occurrenceExceedenceProbability50YearLossTvar": 0.0031,
                "occurrenceExceedenceProbability100YearLossTvar": 0.0063,
                "occurrenceExceedenceProbability500YearLossTvar": 0.017
              }
            },
            {
              "description": "Flood",
              "perilId": "FLOOD",
              "climateScenarioCode": "SSP2_45",
              "averageAnnualLossRatio": 0.00052,
              "averageAnnualLossRiskScore": 14,
              "timeHorizon": 2050,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.0028,
                "annualExceedanceProbability100YearLoss": 0.006,
                "annualExceedanceProbability200YearLoss": 0.0103,
                "annualExceedanceProbability500YearLoss": 0.0168,
                "annualExceedanceProbability50YearLossTvar": 0.0045,
                "annualExceedanceProbability100YearLossTvar": 0.0085,
                "annualExceedanceProbability200YearLossTvar": 0.0141,
                "annualExceedanceProbability500YearLossTvar": 0.0224,
                "occurrenceExceedenceProbability50YearLoss": 0.0023,
                "occurrenceExceedenceProbability100YearLoss": 0.0053,
                "occurrenceExceedenceProbability200YearLoss": 0.0091,
                "occurrenceExceedenceProbability500YearLoss": 0.0148,
                "occurrenceExceedenceProbability50YearLossTvar": 0.0037,
                "occurrenceExceedenceProbability100YearLossTvar": 0.0075,
                "occurrenceExceedenceProbability200YearLossTvar": 0.0124,
                "occurrenceExceedenceProbability500YearLossTvar": 0.0201
              }
            },
            {
              "description": "Wildfire",
              "perilId": "WILDFIRE",
              "climateScenarioCode": "RCP45",
              "averageAnnualLossRatio": 0.0032,
              "averageAnnualLossRiskScore": 68,
              "timeHorizon": 2030,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.015,
                "annualExceedanceProbability100YearLoss": 0.034,
                "annualExceedanceProbability500YearLoss": 0.095,
                "annualExceedanceProbability50YearLossTvar": 0.025,
                "annualExceedanceProbability100YearLossTvar": 0.048,
                "annualExceedanceProbability500YearLossTvar": 0.13,
                "occurrenceExceedenceProbability50YearLoss": 0.013,
                "occurrenceExceedenceProbability100YearLoss": 0.03,
                "occurrenceExceedenceProbability500YearLoss": 0.084,
                "occurrenceExceedenceProbability50YearLossTvar": 0.021,
                "occurrenceExceedenceProbability100YearLossTvar": 0.042,
                "occurrenceExceedenceProbability500YearLossTvar": 0.115
              }
            }
          ]
        }
      ]
    }
  ]
}

Sample response — roof age only

at-get_property_roof_age returns clip and roofAge. There is no climateRisk key.

{
  "count": 1,
  "data": [
    {
      "clip": "123456789",
      "roofAge": [
        {
          "address1": "123 MAIN ST",
          "city": "LOS ANGELES",
          "state": "CA",
          "zipCode": "90001",
          "streetName": "MAIN",
          "latitude": 34.0522,
          "longitude": -118.2437,
          "roofConstructedAge": 16,
          "roofConstructedConfidenceRank": "High",
          "roofConstructedYear": 2010,
          "yearBuilt": 1985,
          "attributes": {
            "medianRoofConstructedAge": 18,
            "medianRoofConstructedYear": 2008,
            "totalNumberOfProperties": 1,
            "yearBuiltModeledIndicator": 0
          }
        }
      ]
    }
  ]
}

Sample response — climate risk only

at-get_property_climate_risk returns clip and climateRisk. There is no roofAge key. The two rows below are the same peril under two scenario-year combinations.

{
  "count": 1,
  "data": [
    {
      "clip": "123456789",
      "climateRisk": [
        {
          "latitude": 34.0522,
          "longitude": -118.2437,
          "craRecordIdentification": "CRA-123456789",
          "modelRunDate": "2025-01-15",
          "modelVersion": "3.0",
          "riskTypeScoreGroup": "SFR",
          "perils": [
            {
              "description": "Wildfire",
              "perilId": "WILDFIRE",
              "climateScenarioCode": "BASE",
              "averageAnnualLossRatio": 0.0029,
              "averageAnnualLossRiskScore": 64,
              "timeHorizon": 0,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.014,
                "annualExceedanceProbability100YearLoss": 0.031,
                "annualExceedanceProbability500YearLoss": 0.088,
                "annualExceedanceProbability50YearLossTvar": 0.023,
                "annualExceedanceProbability100YearLossTvar": 0.044,
                "annualExceedanceProbability500YearLossTvar": 0.121,
                "occurrenceExceedenceProbability50YearLoss": 0.012,
                "occurrenceExceedenceProbability100YearLoss": 0.027,
                "occurrenceExceedenceProbability500YearLoss": 0.078,
                "occurrenceExceedenceProbability50YearLossTvar": 0.019,
                "occurrenceExceedenceProbability100YearLossTvar": 0.039,
                "occurrenceExceedenceProbability500YearLossTvar": 0.107
              }
            },
            {
              "description": "Wildfire",
              "perilId": "WILDFIRE",
              "climateScenarioCode": "RCP85",
              "averageAnnualLossRatio": 0.0041,
              "averageAnnualLossRiskScore": 74,
              "timeHorizon": 2050,
              "attributes": {
                "annualExceedanceProbability50YearLoss": 0.019,
                "annualExceedanceProbability100YearLoss": 0.042,
                "annualExceedanceProbability500YearLoss": 0.118,
                "annualExceedanceProbability50YearLossTvar": 0.031,
                "annualExceedanceProbability100YearLossTvar": 0.059,
                "annualExceedanceProbability500YearLossTvar": 0.161,
                "occurrenceExceedenceProbability50YearLoss": 0.016,
                "occurrenceExceedenceProbability100YearLoss": 0.037,
                "occurrenceExceedenceProbability500YearLoss": 0.104,
                "occurrenceExceedenceProbability50YearLossTvar": 0.026,
                "occurrenceExceedenceProbability100YearLossTvar": 0.052,
                "occurrenceExceedenceProbability500YearLossTvar": 0.143
              }
            }
          ]
        }
      ]
    }
  ]
}

Errors and limits

These tools return the standard MCP ToolError object on failure — code, message, and retryable. Only retry when retryable is true.

Code	Retryable	Typical cause here
`INVALID_INPUT`	No	More than 50 CLIPs in one request, or a CLIP containing non-digit characters
`UPSTREAM_ERROR`	Yes	The upstream data service returned an error
`UPSTREAM_TIMEOUT`	Yes	The upstream data service did not respond in time
`INTERNAL_ERROR`	No	Unexpected server-side failure — escalate if persistent

There is no NOT_FOUND code on these tools. A CLIP with no analytics coverage is returned as a successful response with count of 0 — check count, do not wait for an error.

See MCP tool errors for the full contract and input limits, and Rate limits for 429 handling.

CLIP tools — resolve an address to a CLIP before calling these tools
Property characteristics — structural detail for the same CLIP
Market and trend analytics — market context for the surrounding geography
Response envelope — the data and count shape

​Recommended tool call sequence

​at-get_property_analytics

​Input schema

​Example request

​Output schema

​at-get_property_roof_age

​Input schema

​Example request

​Output schema

​at-get_property_climate_risk

​Input schema

​Example request

​Output schema

​Response envelope

​Response models

​Roof age models

​RoofAgeItem

​RoofAgeAttributes

​Understanding roofConstructedConfidenceRank

​Climate risk models

​ClimateRiskItem

​PerilItem

​Climate scenario codes

​PerilAttributes

​Sample response — full analytics

​Sample response — roof age only

​Sample response — climate risk only

​Errors and limits

​Related

Recommended tool call sequence

`at-get_property_analytics`

Input schema

Example request

Output schema

`at-get_property_roof_age`

Input schema

Example request

Output schema

`at-get_property_climate_risk`

Input schema

Example request

Output schema

Response envelope

Response models

Roof age models

RoofAgeItem

RoofAgeAttributes

Understanding `roofConstructedConfidenceRank`

Climate risk models

ClimateRiskItem

PerilItem

Climate scenario codes

PerilAttributes

Sample response — full analytics

Sample response — roof age only

Sample response — climate risk only

Errors and limits

Related