Taranis MCP for AI. Diagnose crop health from high-res drone imagery.

Listen up. You need a way to run diagnostics that actually work in the field, not some theoretical model. Taranis MCP Server connects your AI agent straight to ultra-high-resolution drone imagery. This thing analyzes everything—weeds, diseases, pests, nutrients—and gives you precise maps telling you exactly what's wrong and where it is.

It’s built for actionable intelligence.

To start, if you need to know what fields are even in play, your agent can use get_organizations to list every farm organization under your account, giving you contact info and a total field count. For a breakdown of the farms themselves, get_fields lists all monitored areas, providing IDs, boundaries, area size, and current crop types for the whole operation.

You also get context about any specific spot with get_field_details, which provides detailed metadata like its exact boundaries, what's planted there, and the general planting context.

When it comes to the raw data feeding this beast, you’ve got several options. To see what flights have happened historically, run get_flights; that gives you all recorded flight dates, weather conditions, resolution specs, and even data quality metrics for a field. If you need the actual visuals, get_flight_imagery pulls high-resolution orthomosaic maps, along with DSMs and DTMs from any specific drone or aircraft mission.

The server also handles specialized imaging layers. For assessing overall crop vigor across the area, get_multispectral_imagery provides essential vegetation indices—specifically NDVI, NDRE, and GNDVI layers—plus statistical summaries for calculation. You can track nutrient health by running get_nutrient_detections, which pinpoints exact locations of nutrient gaps (like nitrogen), estimates how bad the deficiency is, and even suggests variable rate fertilization plans that you'll need.

For a comprehensive view of all problems hitting the farm, your agent uses get_threats. This tool gives a high-level summary across the entire field, including overall severity ratings, what percentage of the area is affected by threats, and trend analysis for the whole season. If you want to map every single crop problem in one go—be it weeds, diseases, pests, or nutrient issues—you'll use get_detections, which returns GPS coordinates along with the severity rating and estimated affected area for all AI-detected crop threats.

When dealing with specific problems, there are dedicated tools. You can pinpoint disease locations using get_disease_detections. This tool retrieves precise spots of crop diseases, identifies pathogen IDs if possible, gives severity levels, and suggests recommended fungicide treatments right away. To hunt down weeds, use get_weed_detections; this pinpoints weed locations, estimates how much area is covered, classifies the species, and offers recommendations for herbicide application.

The server also knows exactly where the money's going to go next. Running get_scouting_recommendations generates AI-powered action plans, suggesting specific patrol routes, priority areas that need checking, and necessary equipment timing for your field team.

This setup gives you a full cycle: you start by gathering inventory data via get_organizations and get_fields, then pull the raw visuals using get_flight_imagery or the spectral layers from get_multispectral_imagery. You run specialized diagnoses—like disease mapping with get_disease_detections or nutrient gap analysis with get_nutrient_detections—and finally, you use get_scouting_recommendations to turn all that data into a concrete plan for the crew.