# Portfolio Image Selection Score AI Agent Connect

> Portfolio Image Selection Score MCP lets your AI client evaluate media libraries against specific project needs. It uses a weighted ranking engine to score assets based on quality, relevance, variety, recency, and rights clearance. Instead of manual sorting, your agent can instantly identify the highest performing images for a specific brief.

## Overview
- **Category:** media-management
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_GbsBPTSofpbQTYIqgc8A8I07sgj8i9jl15IoDWWn/ai-agent-connect
- **Tags:** portfolio, ranking, image-selection, asset-management, scoring

## Description

Managing large media libraries is a headache when you need to find the perfect set of images for a specific client or campaign. This MCP gives your AI agent a mathematical way to judge your assets. You define what matters most for a project—maybe you need high quality and recent shots, or perhaps you need a wide variety of angles—and the MCP calculates a score for every asset in your pool. 

You can control exactly how much weight each factor carries. If legal compliance is your priority, you can weigh rights clearance heavily to ensure you only pick usable content. The system also handles the math of variety, applying penalties if your selection becomes too repetitive. It's a way to turn a massive folder of images into a curated, ranked list that fits your exact requirements in seconds.

## Tools

### calculate_variety_penalty
This tool calculates a penalty score to prevent your asset selection from being too repetitive. It helps your agent pick a diverse set of images.

### filter_by_rights_status
This tool filters your asset list to ensure legal compliance. It isolates only the images that meet your specific rights clearance requirements.

### rank_portfolio_assets
This tool ranks your media assets using a weighted scoring system. It evaluates each item against your specific criteria for quality, relevance, and other metrics.

### validate_weight_distribution
This tool checks your scoring weights to ensure they are mathematically sound. It confirms that all required dimensions are accounted for before you run a ranking.

## Prompt Examples

**Prompt:** 
```
Rank these assets: [{'id': 'img1', 'quality': 0.9, 'relevance': 0.8, 'variety': 0.5, 'recency': 0.7, 'rights_clearance': 1.0}, {'id': 'img2', 'quality': 0.7, 'relevance': 0.9, 'variety': 0.6, 'recency': 0.4, 'rights_clearance': 1.0}] with weights {'quality': 0.4, 'relevance': 0.4, 'variety': 0.1, 'recency': 0.1, 'rights_clearance': 0.0}
```

**Response:** 
```
[{'id': 'img1', 'total_score': 0.81}, {'id': 'img2', 'total_score': 0.76}]
```

**Prompt:** 
```
Check if these weights are valid: {'quality': 0.5, 'relevance': 0.5}
```

**Response:** 
```
{"valid": false, "errors": ["variety, recency, and rights_clearance are missing"]}
```

**Prompt:** 
```
Filter these assets for full rights: [{'id': 'a1', 'rights_clearance': 1.0}, {'id': 'a2', 'rights_clearance': 0.5}]
```

**Response:** 
```
[{'id': 'a1', 'rights_clearance': 1.0}]
```

## Capabilities

### Weighted Asset Ranking
Your agent calculates scores based on custom importance levels for different image attributes.

### Legal Compliance Filtering
The AI isolates assets that meet specific rights clearance thresholds.

### Diversity Enforcement
The system applies penalties to prevent your agent from selecting too many similar images.

### Weight Validation
The MCP verifies that your scoring parameters are complete and accurate.

## Use Cases

### Campaign Asset Selection
An agent picks the best images from a library to match a specific brand's quality and style requirements.

### Rights Management
A user filters a massive collection to find only the images cleared for commercial use.

### Portfolio Curation
The MCP ranks a photographer's work to find the most relevant shots for a new client pitch.

### Content Diversity Checks
The system ensures a social media schedule doesn't use too many similar-looking photos.

## Benefits

- Removes manual sorting by using mathematical scoring for asset selection.
- Ensures legal safety through automated rights status filtering.
- Prevents visual repetition by applying variety penalties.
- Provides consistent results by using a standardized weighting engine.

## How It Works

You connect the MCP to your AI client and start sending asset data for processing.

1. Connect your AI client to the Vinkius-hosted MCP.
2. Provide your list of assets and the weights for each scoring dimension.
3. The agent uses the ranking tool to calculate scores for every asset.
4. The agent applies filters or variety penalties as needed.
5. You receive a ranked, curated list of assets ready for use.

## Frequently Asked Questions

**What dimensions can I use to rank my assets?**
You can rank assets using Quality, Relevance, Variety, Recency, and Rights Clearance.

**How does the MCP handle legal compliance?**
The filter_by_rights_status tool allows your agent to isolate only the assets that meet your specific legal requirements.

**Can I prevent my AI from picking too many similar images?**
Yes, the calculate_variety_penalty tool helps ensure your selection remains diverse.

**Do I need to host this myself?**
No, Vinkius hosts and manages the MCP for you. You just connect your client and start working.

**What happens if my weights don't add up?**
The validate_weight_distribution tool checks your weights to ensure they are mathematically correct and complete.

**How are the assets ranked?**
Assets are ranked by calculating a total score: the sum of each dimension's score multiplied by its assigned weight. In the event of a tie, the asset with the higher individual quality score is ranked higher.

**Can I ensure my weights are valid?**
Yes, you can use the `validate_weight_distribution` tool to verify that your weights are non-negative and sum exactly to 1.0.

**How does the system handle legal usage?**
The `filter_by_rights_status` tool allows you to filter assets based on a hierarchy of clearance levels: full, limited, or none.