Skip to main content
Glama
Sign In
mcp-image-tools

mcp-image-tools

by rog0x
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

mcp-image-tools

MCP server providing image analysis tools for AI agents. Metadata inspection only -- no image processing or manipulation.

Tools

image_metadata

Read image metadata from a URL using HTTP headers. Returns content-type, file size, last-modified, etag, and cache info without downloading the full image.

find_favicons

Find all favicons for any website. Checks /favicon.ico, parses HTML <link> tags, and inspects manifest.json. Returns all discovered favicons with sizes and types.

extract_og_image

Extract Open Graph image, Twitter card image, and Apple touch icon from any URL. Useful for generating link previews.

generate_placeholder

Generate placeholder image URLs via the placehold.co API. Supports custom dimensions, colors, text, format, and font. Returns the URL plus ready-to-use HTML and Markdown markup. Can generate multiple sizes at once.

responsive_images

Generate srcset and <picture> element HTML for responsive images. Given a base image URL, produces multiple size variants with proper markup for responsive design, including multi-format <source> elements.

Related MCP server: simple-vision-mcp

Setup

npm install
npm run build

Usage with Claude Desktop

Add to your Claude Desktop config:

{
  "mcpServers": {
    "image-tools": {
      "command": "node",
      "args": ["path/to/mcp-image-tools/dist/index.js"]
    }
  }
}

License

MIT

Available Tools

5 tools
extract_og_imageA

Extract Open Graph image, Twitter card image, and Apple touch icon from any URL. Returns all social media preview images found on the page.

Inputs: url (required).

find_faviconsA

Find all favicons for a website. Checks /favicon.ico, parses HTML link tags, and inspects manifest.json. Returns all found favicons with sizes and types.

Inputs: url (required).

generate_placeholderA

Generate placeholder image URLs via placehold.co. Supports custom size, colors, text, format, and font. Returns URL plus ready-to-use HTML and Markdown markup.

Inputs: width (required), font, text, sizes, +5 more.

image_metadataA

Read image metadata from a URL via HTTP headers: content-type, file size, last-modified, etag, cache info. No image downloading required.

Inputs: url (required).

responsive_imagesA

Generate srcset and element HTML for responsive images. Given a base image URL, produces multiple size variants with proper HTML markup for responsive design.

Inputs: baseUrl (required), alt, sizes, widths, +3 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    A3.9/5.0

    Scored across 5 tools

    Disambiguation5/5

    Each tool targets a distinct image-related task: reading HTTP headers, finding favicons, extracting social media images, generating responsive markup, and creating placeholders. No two tools could be confused for one another.

    Naming Consistency2/5

    Tool names are a mix of verb-initial names (find_favicons, extract_og_image, generate_placeholder) and noun-phrase names (image_metadata, responsive_images). The lack of a consistent verb_noun pattern makes the set feel less predictable.

    Tool Count5/5

    The server has 5 focused tools, which is well within the ideal range for a single-purpose utility server. Each tool earns its place without bloat.

    Completeness4/5

    The tools comprehensively cover common web image workflows: inspecting, discovering, generating, and integrating images. A minor gap is the lack of image transformation (e.g., resizing) but this is not clearly in scope.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Resources

    Related MCP Connectors

    Related MCP Servers