Image Reader MCP Server

Name: Image Reader MCP Server
Author: moiri-gamboni

@moiri-gamboni

关于 Image Reader MCP Server

An MCP (Model Context Protocol) server (using FastMCP) to provide tools for listing local image files and reading their content. Designed for LLM agent integration.

#mcp #cursor #llm #mcp-server #fastmcp

基本信息

分类

媒体与设计

许可证

GPL-3.0

运行时

node

传输方式

stdio

发布者

moiri-gamboni

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "imageReader": {
      "command": "npx",
      "args": [
        "image-reader-mcp"
      ],
      "env": {}
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Image Reader MCP Server?

A simple MCP server built with FastMCP that allows listing image files in a specified directory and reading a specific image file to return its content as base64 for display. It is intended for users who need to browse and view local image files through an MCP client.

How to use Image Reader MCP Server?

Configure the server in an MCP client by adding an imageReader entry to the mcpServers object, using the command npx with arguments ["image-reader-mcp"]. Invoke the list_images tool with an absolute directory path to see available images, or the read_image tool with the absolute file path of an image to view its content.

Key features of Image Reader MCP Server

Lists image files in a specified directory.
Reads a specific image file and returns its content as base64.
Supports .jpg, .jpeg, .png, .gif, .bmp, .webp, .svg extensions.
Built with FastMCP and uses an imageContent helper for display.

Use cases of Image Reader MCP Server

Browse image files in a local directory via an MCP client.
View a specific image by reading its file content into the client.

FAQ from Image Reader MCP Server

What image formats are supported?

.jpg, .jpeg, .png, .gif, .bmp, .webp, .svg.

Does the server require any runtime dependencies?

It is run via npx, so Node.js and npm are required.

Where does data live?

All image data is read from local files on the user's filesystem; no remote storage is involved.

Are there any known limitations?

When used with Cursor, the server currently seems to function only when Claude Sonnet is selected, as other models don't appear to have vision enabled.