site-intelligence-mcp
# Site Intelligence MCP
<!-- mcp-name: io.github.bayrakdaralper/site-intelligence-mcp -->
Score any geographic coordinate for **ecological and agricultural land suitability**, directly from your AI coding agent.
Analyses satellite and climate data around a point and returns a weighted suitability score with a per-metric breakdown: vegetation health, water access, precipitation, terrain, cropland proximity, urbanisation pressure, climate, and distance to infrastructure.
Built for agtech, precision agriculture, ecological field research, conservation planning, environmental consulting and apiculture.
## Install
Requires a key for the [Site Intelligence API](https://rapidapi.com/bayrakdaralper/api/site-intelligence-api) on RapidAPI (free tier available).
### Claude Code
```bash
claude mcp add site-intelligence \
--env RAPIDAPI_KEY=your_key_here \
-- uvx site-intelligence-mcp
```
### Claude Desktop / Cursor / Windsurf
Add to your MCP config:
```json
{
"mcpServers": {
"site-intelligence": {
"command": "uvx",
"args": ["site-intelligence-mcp"],
"env": {
"RAPIDAPI_KEY": "your_key_here"
}
}
}
}
```
## Usage
Ask your agent in plain language:
> How suitable is 38.42, 27.14 for an apiary? Check May specifically.
> Compare these three coordinates for a reforestation site and tell me which has the best water access.
## Tool
### `site_score`
| Parameter | Type | Default | Notes |
|---|---|---|---|
| `lat` | float | required | -90 to 90 |
| `lon` | float | required | -180 to 180 |
| `radius_m` | int | `2000` | Analysis radius, 200-5000 |
| `month` | int | *(none)* | 1-12 for a seasonal view; omit for annual |
| `detail` | `"summary"` \| `"full"` | `"summary"` | `"full"` adds per-metric raw diagnostics |
`detail` defaults to `"summary"` deliberately: tool results are spent from the
model's context window, and the full diagnostic payload runs to several KB.
Ask for `"full"` only when you need the underlying values.
### Without `uv`
```bash
pip install site-intelligence-mcp
```
then use `site-intelligence-mcp` as the `command` with no `args`.
## Configuration
| Variable | Required | Default |
|---|---|---|
| `RAPIDAPI_KEY` | yes | — |
| `BEELOCATE_RAPIDAPI_HOST` | no | `site-intelligence-api.p.rapidapi.com` |
Get a key from the [Site Intelligence API listing](https://rapidapi.com/bayrakdaralper/api/site-intelligence-api).
## Limits
- Not a weather forecast and not a property-value estimate.
- Resolution is bounded by the underlying satellite sources; very small radii
will not resolve individual parcels.
- Rate limits and quota follow your RapidAPI plan.
## Licence
MIT
TDQS
Scored across 1 tool
With only one tool, there is no possibility of confusion or overlap with other tools. The tool has a clear, singular purpose of scoring a coordinate for suitability.
The single tool name 'site_score' is descriptive and follows a readable noun_verb convention. However, with only one tool there is no established pattern to evaluate consistency across the set.
A single tool feels too thin for an MCP server focused on site intelligence. While the tool itself is substantive, the server's scope would reasonably include at least a few related operations, such as batch comparison or historical analysis.
The server only offers live scoring of one coordinate at a time. Obvious gaps include no way to compare multiple candidate sites, no batch processing, and no persistence or historical trend analysis, which are common needs in site selection and land assessment.