How to set it up

1. 1Open the map and tap the sea. Every layer with data there is listed.
2. 2Press Plan this area in the example project to understand it, simulate the options and read the advice.
3. 3Make your own project to add your data and invite your team.

The address serves every open vector layer: the world ocean layers and the Exmouth Gulf example project’s open data. No sign-in is needed. Each layer states its licence.

1. 1QGIS: Layer → Add Layer → Add WFS / OGC API – Features Layer → New. Paste the address, press OK, then Connect and choose layers.
2. 2ArcGIS Pro: Insert → Connections → Server → New OGC API Server. Paste the address.
3. 3Raster layers (habitat maps, human impact, sea temperature…) come as XYZ tiles. In QGIS: XYZ Tiles → New Connection, with this address. Replace {layer} with a layer id from the Data page.

Not published: a project’s own uploads, cultural data, and data whose owners have not opened it. See every layer the address serves.

1. 1Make an API key below. It acts as you: it sees your projects and the ones shared with you.
2. 2Download the Python client. It needs only requests.
3. 3Look at places, understand areas, run simulations and read your projects, the same way the map does.
4. 4For your own website: call Biohotspot from your server and show the results in your own pages, with your own name and look. Keep the key on the server.

from biohotspot import Biohotspot

bh = Biohotspot("bh_your_key")
here = bh.look(114.45, -22.05, km=5)          # every layer within 5 km
area = bh.understand(lon=114.45, lat=-22.05, radius_km=5)
print(area["car"]["featuresPresent"], "habitats here")
runs = bh.simulate(protect=area["pus"])          # the optimiser, in the example project
seagrass = bh.features("seagrass", bbox=(113.6, -23, 115, -21.4))

The same calls work over plain HTTP, from any language: send Authorization: Bearer bh_… to /engine/api/inspect, /engine/api/assess, /engine/api/marxan/run and /projects/api/projects/.

1. 1Make an API key below.
2. 2Add Biohotspot to your AI tool as an MCP server. The steps for each tool are below. Put your own key where it says bh_your_key.
3. 3Then just ask, for example: what lives near Exmouth, and how much of it is protected? The agent can look at places, check an area, run the options, find layers, search the papers, and read or add data to your projects.

ChatGPT Settings → Apps & Connectors → Advanced settings → turn on Developer mode. Then Create, paste this address, and choose No authentication.

Claude On claude.ai or Claude Desktop: Settings → Connectors → Add custom connector, and paste the same address. In Claude Code, run:

GitHub Copilot In VS Code, add this to .vscode/mcp.json:

{
  "servers": {
    "biohotspot": {
      "type": "http",
      "url": "https://biohotspot.ai4wa.com/mcp",
      "headers": { "Authorization": "Bearer bh_your_key" }
    }
  }
}

Cursor and other agents Most read the same form, under mcpServers instead of servers.

The address with your key in it works like a password. Do not share it. If it gets out, revoke the key below.

Your own data: the agent can add your model’s output to a project from a download link. Biohotspot reads and checks it like any upload, and it appears on the map next to the rest. Only owners and editors of a project can add data.