# History Matching Quality AI Agent Connect

> History Matching Quality MCP lets your AI client quantify the alignment between simulated reservoir production and actual historical observations. It calculates error metrics like RMS and R² to show exactly where your models succeed or fail. Instead of manual data comparison, your agent handles the statistical heavy lifting to identify discrepancies in specific wells or time intervals.

## Overview
- **Category:** simulation
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_zvqfFNstDvVmh4N9etTZx44CAXgYeUdCSq8ylTA7/ai-agent-connect
- **Tags:** reservoir, simulation, history-matching, petroleum, error-metrics

## Description

You can use this MCP to bridge the gap between simulation results and real-world field observations. It provides the mathematical rigor needed to validate reservoir models by comparing simulated production against the observation tier. Instead of squinting at plots, you can ask your AI client to run specific error calculations or statistical comparisons. 

Your agent can pinpoint exactly where a model is drifting. It can isolate specific wells that are driving up your total error or identify specific time windows where the simulation fails to capture reservoir behavior. This makes it much easier to refine your models and ensure they represent the actual field conditions. Whether you need a high-level overview of global match quality or a deep dive into statistical distributions, this MCP provides the direct metrics you need to make informed adjustments to your simulation workflows.

## Tools

### get_feature_comparison_statistics
This tool compares the statistical distributions between simulated and observed data sets.

### get_global_match_metrics
This tool provides a high-level overview of how well the entire simulation matches historical observations.

### get_temporal_error_analysis
This tool detects specific time periods or stages where the simulation fails to capture reservoir behavior.

### get_well_performance_analysis
This tool identifies which specific wells are contributing most to the total error in your simulation.

## Prompt Examples

**Prompt:** 
```
What is the global match quality for my simulation data?
```

**Response:** 
```
The global match quality score is 0.85, with an RMS error of 12.4 and an R² of 0.91.
```

**Prompt:** 
```
Which wells are showing the highest error in the current model?
```

**Response:** 
```
Well W-102 is the primary contributor to error, showing a match quality of 0.42.
```

**Prompt:** 
```
Did the simulation match the production during the first six months?
```

**Response:** 
```
During the interval from 2023-01-01 to 2023-06-30, the R² was 0.88 and the RMS error was 5.2.
```

## Capabilities

### Error Metric Calculation
Your agent calculates RMS, NDS, and R² to quantify model accuracy.

### Well-Level Troubleshooting
The AI identifies specific wells that cause the most significant discrepancies.

### Temporal Error Detection
Your agent finds specific time intervals where the simulation behavior diverges from observations.

### Statistical Distribution Validation
The AI compares the statistical properties of simulated versus observed datasets.

## Use Cases

### Model Validation
Check if a new simulation run actually matches historical production trends.

### Well Troubleshooting
Find out which specific wells are ruining your global match quality.

### Time-Series Analysis
Determine if your model failed to capture a specific pressure drop or production surge during a certain month.

### Statistical Auditing
Compare the distributions of simulated data against observed data to ensure model realism.

## Benefits

- Reduces manual calculation of RMS and R² metrics.
- Isolates problematic wells without manual spreadsheet filtering.
- Detects temporal failures in simulation behavior automatically.
- Provides statistical validation between simulated and observed data.

## How It Works

Connect your AI client to Vinkius and start querying your reservoir data immediately.

1. Connect your preferred MCP-compatible client to Vinkius.
2. Ask your AI client to analyze your simulation data.
3. The AI invokes the specific tool needed for your question.
4. Your agent receives the calculated metrics or analysis.
5. You use the results to refine your reservoir models.

## Frequently Asked Questions

**What metrics does this MCP calculate?**
The MCP calculates RMS error, NDS error, and the R² coefficient of determination.

**Can I find specific wells causing errors?**
Yes, you can use the well performance analysis tool to identify which wells contribute most to the total error.

**How does it handle time-based discrepancies?**
The temporal error analysis tool detects specific time periods where the simulation fails to match reservoir behavior.

**Which AI clients can use this MCP?**
You can use this MCP with any compatible client like Claude, Cursor, Windsurf, or VS Code.

**Do I need to host the MCP myself?**
No, Vinkius hosts and manages the MCP for you, so it is ready to use as soon as you connect.