Command Palette

Search for a command to run...

UnylyUnyly
Весь каталог

Image Analyzer Server

БесплатноНе проверен

Provides local image analysis tools using Pillow for dimensions, colors, metadata, and format conversion without external APIs.

GitHubEmbed

Описание

Provides local image analysis tools using Pillow for dimensions, colors, metadata, and format conversion without external APIs.

README

An MCP (Model Context Protocol) server that provides image analysis tools using Pillow (PIL) — no external APIs required. All processing is done locally.

Features

  • analyze_image — Get dimensions, format, color mode, file size, aspect ratio, and DPI from an image file
  • analyze_image_base64 — Same analysis from a base64-encoded image string
  • get_dominant_colors — Extract dominant colors as hex values with RGB components and approximate coverage percentages
  • get_image_metadata — Read EXIF data including GPS coordinates, camera make/model, date taken, orientation, and more
  • convert_image — Convert between PNG, JPEG, WEBP, and GIF formats, returned as base64

Requirements

  • Python 3.10+
  • mcp>=1.0.0
  • Pillow>=10.0.0

Installation

cd image-analyzer-mcp
pip install -r requirements.txt

Usage

Run with stdio transport (default — for MCP clients)

python server.py

This starts the server over stdio, suitable for integration with MCP clients (e.g., Claude Desktop, Cursor, etc.).

Run with SSE transport (HTTP)

python server.py --transport sse --host 0.0.0.0 --port 8080

Requires uvicorn (pip install uvicorn).

MCP Client Configuration

Add to your MCP client config:

{
  "mcpServers": {
    "image-analyzer": {
      "command": "python",
      "args": ["/path/to/image-analyzer-mcp/server.py"],
      "transport": "stdio"
    }
  }
}

Tool Reference

analyze_image

Analyze an image file.

Parameter Type Required Description
image_path string yes Path to image file

Returns: dimensions, format, mode, file size, aspect ratio, DPI.

analyze_image_base64

Analyze an image from base64-encoded data.

Parameter Type Required Description
base64_data string yes Base64-encoded image data

Returns: Same as analyze_image, but without the file path.

get_dominant_colors

Extract dominant colors from an image.

Parameter Type Required Default Description
image_path string yes Path to image file
count integer no 5 Number of dominant colors (max 128)

Returns: List of dominant colors with hex values, RGB components, and approximate percentage coverage.

get_image_metadata

Extract EXIF metadata from an image.

Parameter Type Required Description
image_path string yes Path to image file

Returns: EXIF data including GPS coordinates, camera info, date taken, orientation, etc.

convert_image

Convert an image to a different format.

Parameter Type Required Default Description
image_path string yes Path to image file
target_format string yes PNG, JPEG, WEBP, or GIF
quality integer no 85 Output quality (1-100, JPEG/WEBP only)

Returns: Base64-encoded converted image data.

License

MIT

Pricing

$19/mo — https://buy.stripe.com/dRm6oJ4Hd2Jugek0wz1oI0m

from github.com/Rumblingb/image-analyzer-mcp

Установка Image Analyzer Server

У этого сервера нет опубликованного пакета — он собирается из исходников. Открой репозиторий и следуй инструкции в README.

▸ github.com/Rumblingb/image-analyzer-mcp

FAQ

Image Analyzer Server MCP бесплатный?

Да, Image Analyzer Server MCP бесплатный — установка в пару кликов через Unyly без оплаты.

Нужен ли API-ключ для Image Analyzer Server?

Нет, Image Analyzer Server работает без API-ключей и переменных окружения.

Image Analyzer Server — hosted или self-hosted?

Self-hosted: сервер запускается локально на твоей машине командой из раздела установки.

Как установить Image Analyzer Server в Claude Desktop, Claude Code или Cursor?

Открой Image Analyzer Server на unyly.org, выбери вкладку своего клиента (Claude Desktop, Claude Code, Cursor) и нажми Install — конфиг сгенерируется автоматически, без правки JSON.

Похожие MCP

Compare Image Analyzer Server with

Не уверен что выбрать?

Найди свой стек за 60 секунд

Автор?

Embed-бейдж для README

Похожее

Все в категории media