# Grain Watch MCP

> Grain Watch connects your AI agent directly to silo temperature monitoring data. Track current conditions, detect hot spots, check humidity levels across multiple silos, and get predictive spoilage risk assessments—all without leaving your chat window.

## Overview
- **Category:** iot-hardware
- **Price:** Free
- **Tags:** silo-monitoring, temperature-sensors, humidity-tracking, predictive-maintenance, spoilage-risk

## Description

Managing a large grain facility used to mean physically checking gauges or staring at dozens of separate dashboard tabs. Now, you keep your AI agent connected via Vinkius's catalog, giving it instant access to the entire silo condition report card. Instead of manually logging readings or waiting for an alarm panel flash, you simply ask your agent what's happening. It instantly pulls current temperature and humidity from every zone in every monitored silo. If a problem is brewing—like a localized hot spot developing in Silo 7—your agent doesn't just tell you; it gives the risk level and recommends immediate next steps. This lets facility managers focus on fixing problems instead of gathering data about them.

## Tools

### get_alerts
Retrieves all active alerts across the facility, detailing critical temperature, humidity, or sensor issues and recommended actions.

### get_current_humidity
Gets current relative humidity percentages from multiple sensors to assess potential condensation risk in a silo.

### get_current_temperature
Pulls real-time temperature readings (in Celsius) from various zones—top, middle, bottom, and center core—within any given silo.

### get_facility_overview
Generates a high-level summary of all monitored silos, detailing general status and overall temperature health across the entire facility.

### get_hotspot_alerts
Detects and reports localized heating events, showing critical hot spots that indicate immediate potential spoilage risk within a silo.

### get_humidity_history
Retrieves time-series data on humidity levels to track moisture migration patterns and detect past condensation events over time.

### get_sensor_health
Checks the operational status, battery life, and communication health of every sensor in a silo's network.

### get_sensor_map
Provides a physical layout map listing where each sensor is placed (depth/zone) within the silo for accurate data interpretation.

### get_silo_details
Pulls specific metadata about a silo, including its grain type and capacity, providing necessary context before analyzing temperature readings.

### get_silos
Lists all monitored silos by name, location, and current status, helping you identify which units need immediate attention.

### get_spoilage_risk
Runs an AI assessment to determine the overall spoilage risk level (low, moderate, high) for a silo based on all collected environmental data.

### get_temperature_history
Generates time-series graphs of temperature readings over a chosen period, helping detect developing hot spots or cooling effectiveness.

## Prompt Examples

**Prompt:** 
```
Show me the current temperature readings for silo 3.
```

**Response:** 
```
Fetching temperature readings for silo 3 (Wheat, 85% full)... Top zone: 18.2C (avg), Middle zone: 19.5C (avg), Bottom zone: 17.8C (avg), Center core: 20.1C. All 24 temperature lines reporting normal. Maximum temperature: 20.1C at center core. Temperature spread: 2.3C (within normal range). Assessment: Grain temperature is NORMAL — no hot spots detected. Temperature gradient is typical for stored wheat. Continue routine monitoring. Would you like to see the 30-day temperature trend?
```

**Prompt:** 
```
Check for any hot spot alerts across my facility.
```

**Response:** 
```
Checking facility hot spot alerts... 1 active hot spot detected. Silo 7 (Corn) — Zone: Center-mid depth, Temperature: 28.5C (7.2C above average). Detected 6 hours ago, trend: Rising. Recommended action: Activate aeration immediately and monitor temperature every 2 hours. All other silos are within normal temperature ranges. Facility average temperature: 19.3C. Would you like me to show the temperature history for silo 7 to see how quickly this hot spot developed?
```

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

**Response:** 
```
Generating facility overview... Total monitored silos: 12. Average temperature: 19.3C (normal range). Active hot spots: 1 (Silo 7, center-mid zone at 28.5C). Temperature alerts: 1 critical, 0 warnings. Sensor health: 96% online (1 sensor offline in Silo 4, scheduled maintenance). Overall spoilage risk: LOW for 10 silos, MODERATE for Silo 7, LOW for Silo 4. Priority action: Address hot spot in Silo 7 with immediate aeration. Schedule sensor replacement in Silo 4 within 1 week. Would you like detailed recommendations for each alert?
```

## Capabilities

### Identify current silo conditions
Get instant temperature readings from all zones, check relative humidity levels, and view the full sensor layout for any specific storage unit.

### Predict spoilage risk
Receive an AI-driven assessment of a silo's overall risk level, including contributing factors and predicted days until failure if conditions don't change.

### Track historical trends
Analyze moisture migration patterns or temperature spikes over time to understand how the grain has been stored previously.

### Monitor facility health
Get a summary of every monitored silo across your entire site, plus a report on which sensors need maintenance or are going offline.

## Use Cases

### Responding to an immediate warning call
An operator gets an alert that Silo 5 is heating up. Instead of manually checking the sensor readings, they ask their agent to run `get_hotspot_alerts` and combine it with `get_sensor_map`. The agent immediately identifies the exact zone and suggests a necessary action.

### Quarterly facility audit
A consultant needs a full report for the client. They use `get_silos` first to list all units, then run `get_facility_overview` to summarize temperature across all 12 silos, generating a comprehensive status report.

### Investigating historical moisture issues
A manager suspects past condensation damage. They ask the agent to pull data from both `get_humidity_history` and then use `get_temperature_history` for the same period, allowing them to see if humidity drops correlated with temperature changes.

### Troubleshooting sensor failures
The system reports an anomaly. The facility manager asks their agent to run `get_sensor_health`. If the report shows a low battery on Sensor 4, they know exactly which piece of equipment needs attention before running any other diagnostic.

## Benefits

- Immediate action on problems. Instead of checking a dashboard for warnings, you can ask your agent to check active hot spots using `get_hotspot_alerts` and get instant alerts about localized heating.
- Understand the full picture. Get facility-wide summaries with `get_facility_overview`, allowing managers to assess overall site health without diving into dozens of individual silo reports.
- Predict failure before it happens. Use `get_spoilage_risk` for an AI assessment that tells you if a silo is at risk and why, shifting your focus from reactive fixing to proactive management.
- Understand the data context. Before analyzing any reading, use `get_silo_details` to verify what type of grain or capacity unit you are actually looking at.
- Track trends over time. Need to know if aeration is working? Use `get_temperature_history` to show a trend line and confirm cooling efforts worked days ago.

## How It Works

The bottom line is, you talk to your AI client like talking to a storage analyst, and it talks back with precise, real-time operational data.

1. Subscribe to the Grain Watch MCP and provide your API key and base URL from your Grain Cloud dashboard.
2. Connect your preferred AI client, like Cursor or Claude, through Vinkius. This establishes the secure link to the monitoring system.
3. Ask your agent a natural language question (e.g., 'Show me all active alerts'). The MCP runs the necessary tool calls and returns actionable data directly into your chat.

## Frequently Asked Questions

**How do I check for active hot spots using the get_hotspot_alerts tool?**
You ask your agent to run `get_hotspot_alerts`. This tool returns the alert severity, affected silo, and temperature differential. It's essential for immediate action because it focuses only on localized heating events.

**What is the difference between get_facility_overview and get_silos?**
`get_silos` just lists all your units by name and location. `get_facility_overview`, however, provides a summary of their operational status and average temperature across the entire site.

**How often should I run get_sensor_health?**
You should check `get_sensor_health` whenever you notice unusual data or when planning maintenance. It tells you which sensors have low batteries, are offline, or need calibration before they fail.

**Can I use get_temperature_history for predictive modeling?**
Yes, `get_temperature_history` provides time-series data that lets your agent build trends. This is critical for detecting developing hot spots or assessing the effectiveness of past cooling efforts.

**What if I need to check humidity in multiple silos?**
You can use `get_current_humidity` and specify multiple silo IDs in your prompt. The agent will then pull relative humidity readings from all specified units for a quick comparison.

**How do I authenticate my connection when running `get_silos`?**
You must provide your API key and base URL in your AI client's configuration. The system uses these credentials to authorize all subsequent calls, like listing available silos or retrieving specific sensor data.

**If I need to check condensation risk, how do I combine `get_temperature_history` with current humidity readings?**
You ask your agent to correlate the two time series datasets. The agent will compare historical temperature spikes against real-time relative humidity levels to accurately pinpoint moisture migration events and potential spoilage windows.

**When running `get_sensor_health`, what should I do if a sensor reports an operational fault?**
An 'operational fault' means the sensor is reading corrupted data, not that it’s offline. You need to physically inspect that specific unit and verify its wiring or calibration against the documentation.

**Can my AI detect hot spots developing in my grain silos before spoilage occurs?**
Yes! Use the `get_hotspot_alerts` tool to check for active hot spot detections across your silos. Hot spots are localized temperature increases that indicate early biological activity (mold, insects, or grain respiration) before visible spoilage. For trend analysis, use `get_temperature_history` to see how temperatures have been changing over the past days or weeks. Early hot spot detection gives you critical time to activate aeration and prevent grain loss.

**How do I get the AI spoilage risk assessment for my silos?**
Use the `get_spoilage_risk` tool with your silo ID. Grain Watch AI analyzes temperature trends, humidity patterns, and grain type to provide a risk level (low, moderate, high, critical), contributing factors, predicted days until spoilage if conditions persist, and recommended preventive actions. This combines multiple data sources into a single actionable assessment. For a facility-wide view, use `get_facility_overview` to see overall risk across all silos.

**Can I check the health of my temperature sensors to ensure reliable monitoring?**
Yes! Use the `get_sensor_health` tool with your silo ID to check the status of all temperature and humidity sensors. This shows which sensors are active, offline, or faulted, along with last communication times and battery levels for wireless sensors. You can also use `get_sensor_map` to see the physical layout of all sensors in a silo, helping you understand which zones each sensor monitors.