# Frame Material Cut List AI Agent Connect

> Frame Material Cut List MCP handles the math for your picture frame production. It calculates exact piece lengths using miter geometry and nests those pieces into your available stock to find the most efficient cutting patterns. Your AI client can now manage production schedules, check material feasibility, and compare different stock lengths to keep waste low.

## Overview
- **Category:** optimization
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_BCbpRjo5JWWGQzexncBl1PIxTgTnSAX2cq0uAfPJ/ai-agent-connect
- **Tags:** cut-list, nesting, framing, material-optimization, production-planning

## Description

Stop manually calculating mitered lengths and guessing how much stock you need for a big order. This MCP handles the geometry and nesting math for you. You give your AI client the artwork dimensions, frame width, and miter allowance, and it returns the exact cut lengths required. 

Beyond simple math, this tool looks at your actual inventory. It calculates how to fit your required pieces into your specific stock lengths to ensure you aren't throwing away expensive material. You can run scenarios to see if a different stock length would be more efficient for a specific job, or quickly verify if a large custom frame is even possible with the material you have on hand. It turns your AI agent into a production planner that understands material constraints and geometric precision.

## Tools

### calculate_cut_list
This tool generates a full production schedule. It provides a list of all required pieces, the stock pieces used, and the remaining offcuts.

### compare_stock_efficiency
This tool evaluates different stock lengths. It tells you which available length results in the least amount of total waste for a specific order.

### get_piece_dimensions
This tool calculates the exact cut length for a single piece. It uses frame geometry to account for mitered corners.

### validate_material_feasibility
This tool checks if your material can actually do the job. It verifies if a set of artwork dimensions can be cut from your current stock.

## Prompt Examples

**Prompt:** 
```
Calculate a cut list for 5 frames. Artwork is 10x12, frame width is 2, miter allowance is 0.5, and stock length is 96.
```

**Response:** 
```
To produce 5 frames, you need 10 pieces of 15.0 units and 10 pieces of 17.0 units. This requires 4 stock pieces of 96 units, with a total of 5 offcuts.
```

**Prompt:** 
```
Is it possible to make a frame for a 20x20 artwork with a 3-inch frame and 0.25 miter allowance using 48-inch stock?
```

**Response:** 
```
No, the required piece length is 26.5 inches, and while a single piece fits, the total material required for all four sides exceeds the capacity of a single 48-inch stock piece.
```

**Prompt:** 
```
Which is better for 10 frames (12x12, 1.5 width, 0.2 allowance): 96-inch stock or 120-inch stock?
```

**Response:** 
```
The 120-inch stock length is more efficient, reducing total waste by 14.5 units compared to the 96-inch option.
```

## Capabilities

### Miter Geometry Math
Your agent calculates exact piece lengths based on frame width and miter allowances.

### Material Nesting
The AI organizes required pieces into available stock lengths to find the best fit.

### Waste Reduction
Your agent compares different stock sizes to find the one that produces the least scrap.

### Feasibility Checks
The AI verifies if specific artwork dimensions can be physically cut from your current inventory.

### Production Scheduling
Your agent generates complete lists of pieces, used stock, and offcuts for the shop floor.

## Use Cases

### Bulk Order Planning
Generate a complete cutting list for dozens of identical frames at once.

### Custom Quote Verification
Check if a unique, large-format artwork can be framed using your current stock lengths.

### Stock Optimization
Compare 96-inch vs 120-inch stock to see which one results in less scrap for a specific job.

### Waste Tracking
Identify exactly how many offcuts you will have left after a production run.

## Benefits

- Reduces manual math errors in mitered corner calculations.
- Lowers material costs by finding the most efficient nesting patterns.
- Speeds up the quoting process by instantly checking material feasibility.
- Provides clear cutting instructions for shop floor workers.

## How It Works

Connect the MCP to your preferred AI client and start calculating.

1. Connect the MCP to Claude, Cursor, or Windsurf via Vinkius.
2. Provide your artwork dimensions and frame specifications to your AI client.
3. Tell the AI to calculate a cut list or check material feasibility.
4. Review the generated cutting schedule or efficiency report.

## Frequently Asked Questions

**How does this MCP handle mitered corners?**
It uses miter geometry to calculate the exact length of each piece based on the frame width and the miter allowance you provide.

**Can I use this with my existing AI tools?**
Yes, you can connect this MCP to any compatible client like Claude, Cursor, or Windsurf through the Vinkius platform.

**Does it help with reducing material waste?**
Yes, it uses nesting logic to fit pieces into stock and includes a tool to compare different stock lengths for maximum efficiency.

**What information do I need to provide for a cut list?**
You typically need the artwork dimensions, frame width, miter allowance, and the length of the stock you are using.

**Can it tell me if a frame is too big for my wood?**
Yes, the validate_material_feasibility tool specifically checks if your dimensions fit within your available stock.

**How does the tool calculate piece lengths?**
The length is calculated by adding the artwork dimension, twice the frame width, and twice the miter allowance.

**Can I compare different stock lengths?**
Yes, you can use `compare_stock_efficiency` to determine which stock length results in the least amount of total waste.

**What happens if a piece is too long for my stock?**
If a piece exceeds the provided stock length, `validate_material_feasibility` will return a failure reason explaining the constraint.
