# Centaur Analytics MCP

> Centaur Analytics connects your AI client directly to industrial grain monitoring systems. You can track CO2, moisture, and temperature trends across multiple bins to predict spoilage risk, manage sensor health, and generate detailed quality reports without leaving your chat interface.

## Overview
- **Category:** iot-hardware
- **Price:** Free
- **Tags:** grain-monitoring, predictive-maintenance, sensor-data, storage-intelligence, spoilage-detection, agritech

## Description

Grain storage is complicated; you can't just look at one reading to know if everything's okay. This MCP lets your AI client take over the job of a full-time grain analyst. Instead of manually checking multiple dashboards for CO2 trends, moisture migration patterns, and hot spots, you talk to your agent and it pulls all that data together. You can get real-time readings for every bin—checking temperature or moisture across different depths in the stored grain. The system also tracks if any sensors are running low on battery or signaling a problem. When you connect this through Vinkius, your AI client acts like a centralized command center, giving you full visibility into the health of your entire facility and predicting exactly when quality might drop. It turns massive amounts of sensor data into simple risk assessments and actionable next steps.

## Tools

### get_facility_overview
Generates an overall report on the entire storage facility's status, useful for management reports that cover everything at once.

### get_alerts
Retrieves immediate notifications about problems like high CO2 or failing sensors, detailing the issue and suggested fixes for specific bins.

### get_bin_details
Pulls basic information about a single storage bin, such as its location, type of grain, and current fill level.

### get_bins
Lists all monitored bins in the facility, providing an inventory count, names, and general monitoring status for overview purposes.

### get_co2_history
Gathers time-series data on CO2 levels to show if biological activity has been slowly increasing or decreasing over a specified period.

### get_current_readings
Provides the instant readings for CO2, moisture, and temperature from all active sensors in a bin right now.

### get_moisture_history
Shows historical moisture content changes to help detect condensation or excessive drying in specific parts of a bin.

### get_quality_forecast
Runs predictive models to estimate what the grain quality will look like weeks from now, helping with marketing timing decisions.

### get_quality_report
Compiles a complete technical report for one bin that combines current data, trends, and actionable recommendations into one document.

### get_sensor_health
Checks the battery life and signal strength of all wireless sensors in a bin to ensure your monitoring system is working properly.

### get_spoilage_predictions
Uses AI models to estimate the specific risk level and the predicted number of days until spoilage starts for a given bin.

### get_temperature_history
Provides historical temperature data, essential for finding 'hot spots' where mold or insects might be active within the grain mass.

## Prompt Examples

**Prompt:** 
```
Show me the current CO2, moisture, and temperature readings for bin 5.
```

**Response:** 
```
Fetching current readings for bin 5 (Corn, 78% full)... CO2: 920 ppm (normal, no biological activity detected). Moisture: 14.1% (stable, within safe storage range). Temperature: Average 12.3C across all sensors, max 13.1C at center, min 11.8C at walls. All 24 wireless sensors reporting healthy. Assessment: Grain condition is EXCELLENT — no spoilage indicators detected. CO2 is well below the 1500 ppm warning threshold. Temperature is uniform with no hot spots. Continue routine monitoring. Next full assessment recommended in 7 days.
```

**Prompt:** 
```
What is the AI spoilage prediction for my soybean bin?
```

**Response:** 
```
Analyzing AI spoilage predictions for soybean bin (bin ID: SB-2847)... Current risk level: LOW. Predicted days until spoilage: 45+ days. Confidence: 87%. Key indicators: CO2 at 780 ppm (stable trend), moisture at 12.8% (well below critical), temperature uniform at 10.5C. AI recommendation: Current storage conditions are optimal. Soybean quality is expected to remain stable through the next 6 weeks. No intervention required. Schedule next quality assessment in 14 days. Marketing recommendation: Current quality supports premium pricing — consider forward contracting if market conditions are favorable.
```

**Prompt:** 
```
Give me a facility-wide overview of all grain bins and any active alerts.
```

**Response:** 
```
Generating facility overview... Total bins: 18. Total inventory: 42,300 tonnes. Average CO2: 980 ppm (normal). Average moisture: 13.9% (stable). Active alerts: 2. Alert 1 (WARNING): Bin 7 — Moisture migration detected at top layer, 15.2%. Recommend targeted aeration. Alert 2 (INFO): Bin 12 — Sensor #8 battery at 15%, schedule replacement within 2 weeks. Overall facility risk: LOW. 16 of 18 bins in excellent condition. Facility quality score: 92/100. Priority actions: Address moisture in Bin 7, replace sensor battery in Bin 12. Would you like detailed recommendations for each alert?
```

## Capabilities

### Assess overall storage condition
Get a snapshot summary of all bins, including their fill levels, grain type, and current monitoring status.

### Monitor real-time sensor metrics
Fetch the immediate CO2 level, moisture percentage, and temperature reading from multiple points within any monitored bin.

### Track historical environmental changes
View time series data for CO2, moisture, or temperature to understand long-term trends like condensation or biological activity increases.

### Predict quality risk and failure dates
Receive machine learning estimates on the likelihood of spoilage and how many days you have before quality degrades.

### Generate formal condition reports
Compile a single, detailed document combining all current data, historical trends, and expert recommendations for documentation or insurance purposes.

## Use Cases

### Detecting a slow problem
A facility manager notices that current readings look fine, but historical data suggests an issue. They ask the agent to run `get_temperature_history` and cross-reference it with `get_co2_history`. The results show gradual temperature creep and rising CO2 trends over two weeks, indicating a slow mold growth problem long before visible damage occurs.

### Preparing for market sale
A commodity trader needs to know the best time to sell. They ask the agent to run `get_quality_forecast` on their soybean bins. The output predicts peak quality metrics three weeks out, allowing them to optimize logistics and timing their sales for maximum profit.

### Responding to a sudden alert
An operator gets an immediate warning about moisture in Bin 7. They use the `get_alerts` tool first, then ask the agent to run `get_bin_details` and compare that info with current readings from `get_current_readings` to determine if aeration or drying is needed immediately.

### Comprehensive reporting
The facility manager needs a report for the bank. They ask for a full assessment, which triggers the agent to use `get_quality_report`. This single action pulls together all necessary sensor data and expert recommendations into one clean document.

## Benefits

- You stop guessing about spoilage. By using `get_spoilage_predictions`, you get a clear risk level and estimate of days until degradation, allowing proactive intervention instead of reactive damage control.
- Facility oversight is instant. Instead of logging into multiple systems, use `get_facility_overview` to see the status of every bin and the overall facility health score in one go.
- You track subtle changes over time. Running `get_co2_history` or `get_moisture_history` reveals trends that simple current readings miss—like slow biological activity or condensation risk.
- You save hours creating paperwork. Generating a full quality assessment using `get_quality_report` compiles every necessary piece of data for compliance or marketing decisions instantly.
- System reliability is guaranteed. Before trusting the data, check `get_sensor_health` to make sure all sensors are reporting and their batteries aren't failing.

## How It Works

The bottom line is that you just talk to your AI client; it does the complex data fetching for you.

1. Subscribe to the Centaur Analytics MCP on Vinkius.
2. Enter your unique API key and base URL credentials into your AI client.
3. Ask your agent a question, like 'What's the spoilage risk in Bin 5?', and it handles the data retrieval.

## Frequently Asked Questions

**How do I check for mold or activity using `get_co2_history`?**
You monitor CO2 levels to find trends. A steadily rising CO2 count is the earliest indicator of biological activity, like mold or insect growth, so keep an eye on that historical data.

**What does `get_spoilage_predictions` tell me?**
It gives you a risk level (low to critical) and estimates how many days the grain quality is expected to remain stable. This helps you plan marketing or storage actions.

**Can I check if my sensors are working with `get_sensor_health`?**
Yes, this tool checks the operational status of every sensor in a bin. It reports battery life and signal strength, so you know which parts of your monitoring system need maintenance.

**How do I get an overall view of all my bins? Use `get_facility_overview`.**
This tool provides a facility-wide summary. It gives total inventory counts, average metrics (CO2, moisture), and flags any active alerts across every bin simultaneously.

**I need to know the current condition: what should I use? Use `get_current_readings`.**
This tool gives you immediate data points—the exact CO2, moisture, and temperature levels right now. It's perfect for a quick check when you walk by the facility.

**How do I get detailed metadata about a specific silo using `get_bin_details`?**
It returns comprehensive context for the bin. You'll find vital details like the grain type, current fill level, and physical location data needed before running any analysis.

**What is the best time to generate a full assessment using `get_quality_report`?**
Run this when you need official documentation for insurance or marketing. It synthesizes current readings, historical trends, and spoilage predictions into one actionable document.

**How do I track moisture changes over time using `get_moisture_history`?**
This function provides time-series moisture data from multiple sensors. You can spot condensation or measure drying effectiveness by observing how the percentage fluctuates across weeks.

**Can my AI predict when grain spoilage will start in my storage bin?**
Yes! Use the `get_spoilage_predictions` tool with your bin ID. Centaur AI analyzes CO2 trends, moisture patterns, and temperature data to predict spoilage risk (low, moderate, high, critical) and estimated days until spoilage onset. For deeper analysis, combine with `get_co2_history` to see the CO2 trend that drives the prediction. CO2 is the earliest spoilage indicator, often rising days before temperature changes become apparent.

**How do I monitor CO2 levels to detect early signs of grain spoilage?**
Use `get_current_readings` for real-time CO2 levels across all sensor positions in a bin, then use `get_co2_history` with a 30-day lookback to identify trends. CO2 levels above 1500 ppm indicate biological activity, and rising trends signal developing spoilage. Set up `get_alerts` to receive automatic warnings when CO2 exceeds safe thresholds. Early CO2 detection gives you 7-14 days more lead time than temperature-based monitoring alone.

**Can I get an AI-generated quality report for a specific bin to share with buyers?**
Yes! Use the `get_quality_report` tool with your bin ID to generate a comprehensive AI-powered quality report. This combines current sensor readings, historical trends, spoilage predictions, and quality forecasts into a single professional report including test weight estimates, moisture stability analysis, temperature uniformity, and mycotoxin risk evaluation. Perfect for buyer communications, insurance documentation, and quality certification.