generated: '2026-08-04'
method: derived
status: candidate
source: Derived from openapi/climateai-weather-openapi.yml. No official ClimateAi MCP server exists. Searched
  climate.ai, docs.climate.ai, the ClimateAI GitHub org (github.com/ClimateAI, 7 public repos — all research/visualization,
  no client or server tooling), npm, and PyPI on 2026-08-04; nothing published. ClimateAi does market
  an "AI agents" webinar (https://climate.ai/downloadable-content/webinar-ai-agents/) describing agent-driven
  monitoring on its own platform, but ships no agent-facing protocol surface.
server:
  name: climateai-weather
  transport: http
  url: null
  auth: apiKey header X-Api-Key — see authentication/climateai-authentication.yml
notes: 'A candidate MCP tool projection over the LensConnect Weather API. Each tool maps 1:1 to a real
  operationId in the published OpenAPI, so each tool''s real inputSchema is that operation''s `parameters`
  block — no input contract is invented here. Only the weather surface is projected: the ClimateAI Platform
  gateway is an internal tenancy and routing API (accounts, users, roles, device keys) with destructive
  writes and no agent use case, so it is deliberately excluded from the candidate tool set. status: candidate
  means ClimateAi does not operate this server; it is a design proposal, not a claim about the provider.'
tools:
- name: get_history
  description: Historical weather for a coordinate, 1995-present (ERA5-backed), with variable selection,
    date range, daily/weekly/monthly aggregation and optional ~1 km downscaling.
  source_operation: openapi/climateai-weather-openapi.yml#getHistoryGrid
  inputs:
  - lat
  - lon
  - var
  - start_date
  - end_date
  - granularity
  - use_downscaling
- name: get_forecast_ensemble
  description: Stitched forecast from today through ~6 months as raw ensemble members (~31 values per
    date) for computing custom risk metrics.
  source_operation: openapi/climateai-weather-openapi.yml#getStitchedForecast
  inputs:
  - lat
  - lon
  - var
  - start_date
  - end_date
  - use_downscaling
- name: get_forecast_statistics
  description: Stitched forecast as mean plus configurable quantiles, aggregated daily, weekly or monthly.
    The default agent-facing forecast tool.
  source_operation: openapi/climateai-weather-openapi.yml#getStitchedForecastStatistics
  inputs:
  - lat
  - lon
  - var
  - start_date
  - end_date
  - use_downscaling
  - statistics
  - granularity
- name: get_climatology
  description: Long-term statistical baseline for a coordinate over the fixed 1993-2022 reference period,
    at daily, weekly or monthly granularity.
  source_operation: openapi/climateai-weather-openapi.yml#getClimatology
  inputs:
  - lat
  - lon
  - var
  - granularity
- name: get_climatology_by_location
  description: Climatology for a stored server-side location ID.
  source_operation: openapi/climateai-weather-openapi.yml#getClimatologyByLocation
  inputs:
  - id
  - var
  - granularity
- name: get_history_legacy
  description: Legacy rolling ~366 days of daily observations for a coordinate.
  source_operation: openapi/climateai-weather-openapi.yml#getHistory
  inputs:
  - lat
  - lon
  - var
- name: get_history_by_location
  description: Legacy history for a stored server-side location ID.
  source_operation: openapi/climateai-weather-openapi.yml#getHistoryByLocation
  inputs:
  - id
  - var
- name: get_short_term_forecast
  description: Legacy short-term (~15 day) daily forecast for a coordinate.
  source_operation: openapi/climateai-weather-openapi.yml#getShortTermForecast
  inputs:
  - lat
  - lon
  - var
- name: get_subseasonal_forecast
  description: Legacy subseasonal (weeks 2-6) forecast. Weekly responses add probabilistic terciles that
    no v2 endpoint exposes.
  source_operation: openapi/climateai-weather-openapi.yml#getSubseasonalForecast
  inputs:
  - lat
  - lon
  - var
  - granularity
- name: get_seasonal_forecast
  description: Legacy seasonal (months 1-6) forecast. Monthly responses add probabilistic terciles that
    no v2 endpoint exposes.
  source_operation: openapi/climateai-weather-openapi.yml#getSeasonalForecast
  inputs:
  - lat
  - lon
  - var
  - granularity
- name: get_short_term_forecast_by_location
  description: Legacy short-term forecast for a stored server-side location ID.
  source_operation: openapi/climateai-weather-openapi.yml#getShortTermForecastByLocation
  inputs:
  - id
  - var
- name: get_subseasonal_forecast_by_location
  description: Legacy subseasonal forecast for a stored server-side location ID.
  source_operation: openapi/climateai-weather-openapi.yml#getSubseasonalForecastByLocation
  inputs:
  - id
  - var
  - granularity
- name: get_seasonal_forecast_by_location
  description: Legacy seasonal forecast for a stored server-side location ID.
  source_operation: openapi/climateai-weather-openapi.yml#getSeasonalForecastByLocation
  inputs:
  - id
  - var
  - granularity
agent_notes:
- Every tool is a safe read; there is no destructive operation in this projection.
- Resolution is entitlement-dependent. use_downscaling=true does nothing unless the key is provisioned
  for 1 km, so an agent must read the response meta rather than assume.
- No rate-limit contract is published, so an agent has no budget signal and should apply its own conservative
  backoff.
- see: mcp/climateai-tool-crosswalk.yml
deployment:
  mode: none
  verified: derived
  tools: 13
  checked: '2026-08-12'
  source: catalog MCP census