# Taranis MCP

> Taranis connects AI agents directly to ultra-high-resolution drone imagery for advanced crop diagnostics. It analyzes field data to pinpoint weeds, diseases, pests, and nutrient deficiencies across entire fields. Your agent receives detailed maps showing threat locations (GPS coordinates), severity levels, affected areas, and specific action plans needed for targeted scouting or treatment.

## Overview
- **Category:** the-unthinkable
- **Price:** Free
- **Tags:** precision-agriculture, crop-scouting, imagery-analysis, threat-detection, field-management, computer-vision

## Description

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.

## Tools

### get_fields
Lists all monitored farm fields, providing IDs, boundaries, area size, and current crop types for an organization.

### get_clients
Returns client names, IDs, and associated farm counts. Use this after get_organizations to navigate the hierarchy: Organizations → Clients → Farms → Fields.

List clients within a specific Taranis organization

### get_detections
Returns GPS coordinates and severity for all AI-detected crop threats, including weeds, diseases, pests, and nutrients.

### get_disease_detections
Retrieves locations of crop diseases, pathogen IDs, severity levels, and suggested fungicide treatments.

### get_farms
Returns farm names, IDs, locations, and field counts. Use this after get_clients to navigate to specific farms before querying fields.

List farms belonging to a specific client

### get_field_details
Gives detailed metadata about a specific field, including its boundaries, crop type, and planting context.

### get_map_layers
Returns layer metadata and download URLs. Essential for crop vigor assessment, variable rate application planning, and growth stage monitoring.

Get map layers (NDVI, imagery, multispectral) for a specific field

### get_nutrient_detections
Identifies nutrient deficiency locations, estimates severity, and suggests variable rate fertilization plans.

### get_organizations
Lists all farm organizations under your account with basic contact info and total field counts.

### get_scouting_recommendations
Generates AI-powered action plans, suggesting specific patrol routes, priority areas, and necessary equipment for the field team.

### get_threats
Provides a consolidated summary of all threats, including overall severity ratings, affected percentages, and trend analysis per field.

### get_weed_detections
Pinpoints weed locations, estimates coverage area, classifies species, and gives recommendations for herbicide use.

## Prompt Examples

**Prompt:** 
```
Show me all AI-detected threats in my corn field from the latest flight.
```

**Response:** 
```
Analyzing latest flight detections for corn field (field ID: CF-2847)... Found 23 detections across 4 threat categories. WEEDS: 12 patches detected — Waterhemp (7 patches, 2.3% coverage), Palmer Amaranth (3 patches, 1.1%), Giant Ragweed (2 patches, 0.8%). DISEASES: 6 detections — Gray Leaf Spot (moderate severity, 4 locations), Northern Corn Leaf Blight (early stage, 2 locations). PESTS: 3 detections — Corn Rootworm adults (2 locations), European Corn Borer damage (1 location). NUTRIENTS: 2 nitrogen deficiency zones (1.5% combined area). Overall field health: Good with localized treatment needed. Top priority: Waterhemp patches in northeast quadrant. Would you like scouting recommendations for these detections?
```

**Prompt:** 
```
Generate scouting recommendations for my soybean field this week.
```

**Response:** 
```
Generating AI scouting recommendations for soybean field (field ID: SB-1923)... Priority 1 (HIGH): Verify Sudden Death Syndrome symptoms in southeast corner — 4 disease detections with moderate severity. Walk pattern: Entry at gate 3, follow east boundary south, check 4 GPS-marked locations. Priority 2 (MEDIUM): Scout for soybean aphids in north-central area — NDVI shows stress patches consistent with early aphid pressure. Priority 3 (LOW): Check nitrogen deficiency zones in western section — 3 small patches identified, consider in-season N application if economic threshold met. Recommended scouting time: 45-60 minutes. Best conditions: Morning hours, dry foliage. Would you like GPS coordinates for each scouting location?
```

**Prompt:** 
```
What is the overall threat level and NDVI trend for my wheat field this season?
```

**Response:** 
```
Analyzing wheat field (field ID: WH-5512)... Current threat level: MODERATE. Season summary: 3 flights completed, 18 total detections. Disease trend: Septoria tritici increasing from 2 to 6 detections over 3 flights — monitor closely. Weed trend: Annual ryegrass stable at 5 patches (1.8% coverage). Nutrient trend: Nitrogen deficiency decreasing after top-dress application — 3 zones down to 1 zone. NDVI trend: Flight 1 (0.45) — tillering stage, Flight 2 (0.68) — stem extension, Flight 3 (0.82) — current heading stage. Field performing above regional average. Recommendation: Continue disease monitoring, no immediate treatment required. Next flight scheduled in 10 days.
```

## Capabilities

### Map all crop threats
The `get_detections` tool returns GPS coordinates for weeds, diseases, pests, or nutrient issues, along with severity and estimated affected area.

### Diagnose specific crop diseases
Use `get_disease_detections` to pinpoint disease locations, identify pathogens where possible, and receive suggested fungicide treatments.

### Analyze nutrient deficiencies
`get_nutrient_detections` identifies specific nutrient gaps (like nitrogen) and suggests variable rate fertilization plans for the field.

### Assess overall field health status
The `get_threats` tool provides a high-level summary, including overall severity ratings and trend analysis across the entire monitored area.

### Generate targeted scouting plans
Run `get_scouting_recommendations` to receive prioritized action items, specific patrol patterns, and recommended timing for field teams.

### Retrieve multispectral indices
The `get_multispectral_imagery` tool provides NDVI, NDRE, and GNDVI maps essential for calculating crop vigor across the field.

## Use Cases

### Post-flight damage assessment
A farm manager needs to know what happened after a major storm. They ask their agent for the overall status. The agent first uses `get_flights` to find the latest images, then runs `get_threats` and `get_detections`. The result is a consolidated report showing all current threats—pests, weeds, etc.—without manual data comparison.

### Planning for maximum yield
An agronomist suspects the crop isn't performing well. They run `get_multispectral_imagery` to check NDVI and then use `get_nutrient_detections`. The agent combines these, showing a correlation between low vigor (low NDVI) and specific nitrogen deficiency zones, leading directly to a treatment recommendation.

### Responding to an outbreak
A field team reports unusual spots. The consultant uses `get_disease_detections` immediately to map the pathogen spread and severity. They then use `get_field_details` to confirm the crop type, ensuring the fungicide recommendation is accurate for that specific plant.

### Creating a preventative maintenance schedule
A consultant needs to advise a client on future work. The agent first uses `get_fields` to list all sites and then runs `get_scouting_recommendations`, giving the client an immediate, prioritized checklist of action items for every single field.

## Benefits

- Targeted scouting becomes efficient. Instead of guessing, let your agent run `get_scouting_recommendations` to get a prioritized patrol route and specific GPS locations for the field team.
- You stop comparing spreadsheets. Running `get_multispectral_imagery` immediately provides NDVI/NDRE maps, letting you assess crop vigor across an entire field in minutes.
- Know exactly where the money is needed. By calling `get_nutrient_detections`, your agent doesn't just say 'Nitrogen is low'; it gives variable rate fertilization plans for specific zones.
- Get a single health score. The `get_threats` tool summarizes everything—weeds, diseases, nutrients—into one overall severity rating and trend analysis per field.
- Pinpoint the problem quickly. If you suspect disease, running `get_disease_detections` immediately returns pathogen IDs and suggested fungicides, cutting diagnosis time by hours.

## How It Works

The bottom line is you stop manually opening 12 different dashboards and start asking your agent for a single, combined diagnostic report.

1. First, run `get_fields` or `get_organizations` to define the exact scope (field ID) of your analysis.
2. Next, chain multiple detection tools—like running both `get_weed_detections` and `get_disease_detections`—to gather all relevant data layers for that field.
3. Finally, ask your agent to summarize the results using `get_scouting_recommendations`, forcing it to combine those disparate reports into one action plan.

## Frequently Asked Questions

**How do I start analyzing threats in Taranis MCP Server?**
You must first call `get_fields` to confirm the field ID. Once you have that ID, you can then run tools like `get_detections` or `get_threats` using the returned identifier.

**Can I check for multiple types of problems at once with Taranis MCP Server?**
Yes. You combine calls to different detection endpoints, such as running both `get_weed_detections` and `get_disease_detections` in a single agent workflow for comprehensive coverage.

**What is the difference between using get_threats and get_detections?**
`get_detections` gives you the raw, specific data (GPS coordinates, severity) for each threat type. `get_threats` summarizes that data into a high-level status report with overall trend analysis.

**Does Taranis MCP Server handle historical imagery?**
Yes. Use `get_flights` to list past missions, and then use `get_flight_imagery` or `get_multispectral_imagery` to access the specific maps from those dates.

**What should I do if my field has nutrient deficiencies?**
Run `get_nutrient_detections`. The tool provides deficiency locations and, critically, suggests variable rate fertilization plans so you know exactly what product goes where.

**How do I ensure my AI agent has access to all farm data using get_organizations?**
You must call `get_organizations` first. This lists every organization ID you have access to. You need this list to properly scope subsequent calls, like running a field check or detection analysis across multiple sites.

**What specific indices are available when I use get_multispectral_imagery?**
It provides key vegetation indices including NDVI (Normalized Difference Vegetation Index), NDRE, and GNDVI. These layers help assess crop vigor levels beyond simple visual checks. The data also includes statistical summaries like mean and standard deviation.

**Why is it important to run get_field_details before analysis?**
It establishes the necessary context for any work. This tool returns critical metadata, including boundaries in GeoJSON format, crop type, and growth stage. Running this first ensures your AI agent knows exactly what it's looking at.

**Can my AI detect specific weed species in my soybean field from Taranis imagery?**
Yes! Use the `get_weed_detections` tool with your field ID to get AI-detected weed infestations with species-level identification. Returns weed locations, estimated coverage area, species classification, growth stage, and herbicide resistance indicators. For a comprehensive view of all threats (weeds, diseases, pests, nutrients), use `get_detections` without a type filter.

**How do I get scouting recommendations based on the latest flight imagery?**
Use the `get_scouting_recommendations` tool with your field ID. Taranis AI analyzes the latest imagery, detected threats, crop growth stage, and field history to generate specific action items including ground truth verification locations, recommended scouting patterns, treatment suggestions, and priority levels. You can also use `get_threats` first to see the overall threat severity before reviewing recommendations.

**What resolution imagery does Taranis capture and how often are flights conducted?**
Taranis captures ultra-high-resolution imagery at sub-millimeter to centimeter level resolution using specialized drone and fixed-wing aircraft. Flight frequency depends on your monitoring plan and crop growth stage — typically every 7-14 days during critical growth periods. Use `get_flights` to see all flight history for a field, and `get_flight_imagery` to access specific imagery products (orthomosaics, DSM, NDVI maps) from any flight mission.