Agriculture MCP Server

Full-stack agriculture MCP server with 35 tools — farm management, weather intelligence, satellite imagery, commodity markets, IoT sensors, and yield forecasting. Includes a built-in example AgTech backend with realistic Kenyan farm data.

Quick Start

# 1. Start the example backend (seeded with Kenyan farm data)
cd example-backend && cargo run &

# 2. Run the MCP server (connects to localhost:7800)
cd .. && cargo run

No API keys needed — the example backend provides all intelligence data locally.

Tools (35)

| Category | Tools | Description | |----------|-------|-------------| | Fields (5) | list, get, create, update, delete | Farm plots with GPS, soil type, elevation | | Crops (5) | list, get, plant, calendar, log_harvest | Crop lifecycle from planting to harvest | | Activities (3) | log, list, summary | Farm operations (irrigation, fertilizer, spraying) | | Weather (4) | forecast, historical, alerts, GDD | 7-day forecast, growing degree days | | Satellite (4) | NDVI, crop_health, anomalies, boundary | Remote sensing crop assessment | | Market (5) | price, history, trends, commodities, best_sell_time | Commodity prices and sell recommendations | | IoT (4) | sensors, reading, soil_moisture, rainfall | Real-time sensor data | | Yield/Alerts (5) | estimate, harvest_history, compare_seasons, pests, disease_risk | Intelligence and alerts |

Example Backend

The example-backend/ is a standalone Axum server with SQLite, seeded with:

• 5 fields: Kiambu Coffee Estate (12.5ha), Nakuru Wheat (45ha), Meru Tea (8.2ha), Machakos Maize (5ha), Kericho Avocado (3.5ha)
• 5 crops: SL28 Coffee, Kenya Eagle Wheat, TRFK 6/8 Tea, H614D Maize, Hass Avocado
• 5 IoT sensors: Soil moisture, temperature, rain gauge, humidity
• Intelligence: Weather forecasts, NDVI readings, commodity prices (KES), pest alerts, disease risk

All data is realistic for Central/Western Kenya growing conditions.

Configuration

With example backend (zero config)

export AGTECH_API_URL=http://localhost:7800/api/v1

With external AgTech platform

export AGTECH_API_URL=https://your-platform.com/api/v1
export AGTECH_API_KEY=your-key

Client Configuration

{
  "mcpServers": {
    "agriculture": {
      "command": "mcp-agriculture",
      "args": [],
      "env": {
        "AGTECH_API_URL": "http://localhost:7800/api/v1"
      }
    }
  }
}

Usage Examples

"List all my farm fields"
→ list_fields()

"What's the weather forecast for my coffee field?"
→ get_forecast(field_id: "f1")

"How healthy is my coffee crop?"
→ get_crop_health(field_id: "f1")

"What's the current coffee price?"
→ get_commodity_price(commodity: "coffee")

"Should I sell my coffee now or wait?"
→ get_best_sell_time(commodity: "coffee")

"What's the soil moisture in my maize plot?"
→ get_soil_moisture(field_id: "f4")

"Any pest alerts for my region?"
→ get_pest_alerts()

"Estimate my coffee yield this season"
→ estimate_yield(field_id: "f1")

Registry Compliance

• HealthCheck — verifies AgTech backend connectivity
• mcp-server.toml — 35 tools with risk classes
• Manifest validation — startup fails fast on invalid manifest

License

Apache-2.0

Part of the ADK-Rust Enterprise MCP server ecosystem.