Criora climate risk

Climate and disaster risk for any place: 7-day weather risk, hazards, climate profile, country risk.

CommunityOnlineNo sign-inGlobalFreeRead-only

What it can do

    What data it sees

    Do you need an account

    No: the server works without sign-in

    Climate and disaster risk for any place: 7-day weather risk, hazards, climate profile, country risk.

    How to connect

    How to connect Criora climate risk to Claude

    The server address to paste into Claude settings:

    https://criora.com/mcp

    1. Open Claude (claude.ai or the desktop app).
    2. Go to Settings → Connectors.
    3. Click Add custom connector.
    4. Paste the server address copied below into Remote MCP server URL and click Add.
    5. In a chat, click + → Connectors and switch the new connector on.

    Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.

    Authorization

    After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.

    How to connect Criora climate risk to ChatGPT

    The server address to paste into ChatGPT settings:

    https://criora.com/mcp

    1. Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
    2. Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
    3. Open Settings → Connectors and click Create.
    4. Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
    5. Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
    6. In a new chat open + → Apps/Connectors and enable the connector.

    OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.

    Authorization

    On first use ChatGPT opens the service's sign-in window. Sign in and approve access.

    How to connect Criora climate risk to Cursor

    The server address to paste into Cursor settings:

    https://criora.com/mcp

    Fastest: click Open in Cursor below and confirm the prompt.

    Manually:

    1. In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
    2. Paste the JSON copied below into ~/.cursor/mcp.json (per project: .cursor/mcp.json in the repo root).
    3. Save the file. The server appears in the MCP list; authorize it there if asked.

    Authorization

    If the service needs sign-in, an authorize button appears next to the server in the MCP list.

    Server tool list (12)

    Raw names from tools/list. Only developers need these.

    show_risk_mapShows an interactive map of a place in clients that render apps, such as Claude and ChatGPT: this week's forecast risk over the area, hazard events nearby as markers, and the risk by day, which the reader can pan, zoom and open full screen. Instead of the risk, the map can lay the precipitation, temperature or wind forecast under the events (`layer`), hour by hour over the week, and the reader can switch between all four. A client without apps gets the same summary as text, with a link to the place on Criora's own map. Use it when the user wants to see the place, the week's risk or the hazards around it on a map, or when a map makes an answer clearer than a list. The figures come from the same sources as get_forecast and get_hazards_near, which give them in full. Give latitude and longitude, or place, which costs a lookup in the shared geocoder. With both, the coordinates are used and place only names the map, at no lookup: pass the name the user knows the place by, or the map is titled by its coordinates and country.
    get_map_eventsThe hazard events the map draws for a view the reader has moved to. Called by the map, never shown to the model, which has get_hazards_near.
    find_placeCandidate coordinates for a place name or address, best match first, each with its full name and kind (city, state, road and so on). Call it when a name is ambiguous (several Springfields) and the right candidate must be chosen, or when one place is the subject of several calls and its coordinates should be looked up once. The other place tools take a name themselves and use its best match, so a single question about an unambiguous place needs no find_place first. It queries OpenStreetMap Nominatim (ODbL) through Criora's geocoder, which every user shares and which sends one request a second: a burst of calls waits its turn, and a name looked up in the last week is answered from a cache. A name that matches nothing returns an empty list. A city resolves to its centre as OpenStreetMap maps it.
    get_forecastThis week's forecast at a place by local day: minimum and maximum temperature, peak precipitation rate, peak wind, and the day's forecast risk class with the hazard behind it (heat, cold, wind, rain, snow, freezing rain) and its physical reading (UTCI felt temperature, gust speed or rain rate). Use it for the weather or the weather risk in the next seven days: it answers that at a third of the size of assess_place, with every step on request. For events already happening (fires, floods, earthquakes) use get_hazards_near; for what a place is exposed to over years, get_climate_profile. Give latitude and longitude, or place, which costs a lookup in the shared geocoder; with both, the coordinates are used. hourly=true adds every forecast step to the daily summary. Steps are hourly for the first day, then 3-hourly, then 6-hourly to day seven, and each day gives its step_hours: from the fourth day the minimum and maximum are of 6-hourly steps, so the true peak can fall between two of them. A day marked partial is cut by the start or end of the window. Source: Environment and Climate Change Canada GDPS global model, read at its own 15 km cell containing the point.
    get_climate_profileLong-term climate exposure at a place: every climate layer on the Criora map, read at the point. ERA5 extremes of the latest full year, CMIP6 warming and rainfall change to 2050 and 2080, flood depth by scenario, sea level rise, water stress, drought, aridity, burned area, land cover, elevation, subsidence and the Köppen zone. Each layer gives its year, unit, cell size and readings by band, and names its source, whose institution, licence and link are given once in sources. Use it for what a place is exposed to over years, and once per place to compare places. It has every layer where assess_place has the headline ones, and is larger for it, so pass categories when the question is about one kind of exposure. For this week's weather use get_forecast; for events happening now, get_hazards_near. Give latitude and longitude, or place, which costs a lookup in the shared geocoder. With both, the coordinates are used and place only names the map, at no lookup: pass the name the user knows the place by, or the map is titled by its coordinates and country.
    get_hazards_nearDisasters and hazards happening now or recently near a place, most severe and most recent first, each with its distance, bearing, severity and time. Wildfires (NASA FIRMS), earthquakes, floods, tropical cyclones, volcanic eruptions and droughts (GDACS), US National Weather Service flood, tropical cyclone and fire warnings, and air quality (OpenAQ reference monitors). Use it for what is happening there now: it answers that at a fraction of the size of assess_place, and the radius, event types and period can be chosen. For the weather forecast use get_forecast; for long-term exposure, get_climate_profile. Give latitude and longitude, or place, which costs a lookup in the shared geocoder; with both, the coordinates are used. Without radius_km every event type is searched at the largest distance, which radius_used_km reports. The most severe and most recent events are listed, and total and counts_by_type cover every event found.
    get_country_profileA country's climate and disaster risk from the Criora Atlas: overall risk score, class and rank among all countries, the climate, nature, social, governance and adaptation scores and classes, and its income group and region. Use it for one country in depth. For several named countries side by side use compare_countries; to find which countries are most or least at risk, rank_countries; for a place inside a country, the place tools. The country is an ISO 3166-1 alpha-3 code or its English name; an unknown one is refused with the expected form. Scores run 0 to 100; higher means more risk. include_indicators=true adds, per dimension, the indicators where the country is furthest from the world median, first (INFORM Risk Index, ND-GAIN, World Bank), each with its value, year, source and the medians of the world and of the country's own region, subregion and income group, the ones meta names. indicators_limit sets how many are listed per dimension, and total counts each dimension's indicators.
    compare_countriesTwo to ten named countries side by side: each one's overall risk score and class, its climate, nature, social, governance and adaptation scores, and its rank by overall risk. Use it when the countries are known in advance. To find the countries most or least at risk, optionally within a region or income group, use rank_countries; for one country's underlying indicators, get_country_profile. Countries are ISO 3166-1 alpha-3 codes or English names; an unknown one is refused with the expected form. Scores run 0 to 100; higher means more risk. Ranks are among all countries in the Criora Atlas, 1 being the highest risk.
    rank_countriesThe countries most or least at risk, overall or on one risk dimension, each with its score, risk class and rank, optionally only those in a region, subregion or income group. Answers "which countries are most exposed to climate risk", "the riskiest countries in Southern Asia" or "low-income countries with the lowest adaptation score". For countries named in advance use compare_countries; for one country in depth, get_country_profile. Regions, subregions and income groups are named as the countries' meta names them; a name no country has is refused with the names there are. The filters combine, and a combination that matches nothing returns an empty list; limit cuts the list, while matched counts every country that passed the filters. Scores run 0 to 100; higher means more risk. Each country's rank is among all countries in the Criora Atlas, 1 being the highest risk: a filter narrows the list but not the rank. Countries without a score on the dimension are not ranked and are counted in no_score.
    assess_placeOne answer for an open question about one place ("is it safe there", "what are the risks"): this week's forecast risk by day, hazard events nearby, the headline climate exposures (heat, flood, sea-level rise, water stress, drought, heavy rain, fire, climate zone) and the risk profile of the country it lies in. For a specific question, call the single tool instead, which answers it in full: this week's weather, get_forecast; events happening now, get_hazards_near; long-term exposure, get_climate_profile, which has every climate layer where this has the headline ones; the country, get_country_profile. It answers for one place: places are compared on get_climate_profile, one call each, layer by layer, and countries on compare_countries or rank_countries. Give latitude and longitude, or place, which costs a lookup in the shared geocoder. With both, the coordinates are used and place only names the map, at no lookup: pass the name the user knows the place by, or the map is titled by its coordinates and country.
    searchCriora documents matching a query, each with the id fetch reads and the Criora page to cite: countries, places, and the pages on Criora's methodology and data sources. Use it to find what Criora has on a subject before reading it with fetch. For an answer in one call, the specific tools are shorter: assess_place for a place, get_country_profile for a country, rank_countries for the countries most at risk. A country is found by its ISO 3166-1 alpha-3 or alpha-2 code or its English name as the whole query, or by its name anywhere in it. The words methodology or sources find those pages. Only when nothing else matches is the query looked up as a place name, once, in the geocoder every user shares, which answers one lookup a second; a place is found best by its name alone. Nothing found is an empty list.
    fetchOne Criora document as markdown, with its title, the Criora page to cite and the sources it draws on, by an id search returned. country:<ISO3> (country:BGD) gives the country's overall risk score, class and rank, each dimension with its class, its most notable indicators against the world and regional medians, and the sources with their dates. place:<latitude>,<longitude> (place:23.8103,90.4125) gives this week's forecast risk by day, the five most notable hazard events nearby, the headline climate exposures and the country. page:methodology and page:sources are those pages of the site. Use it to read a document for research and cite it. The specific tools answer with every figure as data: get_country_profile, assess_place and the rest. An id of another form is refused with the forms it takes. A long text is cut at a paragraph, with the page that has the rest.