MCP Server
Connect Claude, GPT, Cursor, or any MCP client to the Spronta Image CDN. Upload images, create presets, generate signed URLs — all through natural language.
Claude Code
claude mcp add spronta \
-e SPRONTA_API_KEY=spronta_img_... \
-e SPRONTA_PROJECT_ID=your-project-id \
-- npx @spronta/mcpClaude Desktop / Cursor
{
"mcpServers": {
"spronta": {
"command": "npx",
"args": ["@spronta/mcp"],
"env": {
"SPRONTA_API_KEY": "spronta_img_...",
"SPRONTA_PROJECT_ID": "your-project-id"
}
}
}
}The Spronta MCP Server is an open-source tool that connects AI assistants to the Spronta Image CDN using the Model Context Protocol. It exposes 18 tools that let LLMs manage image projects, upload and optimize images, create transform presets, generate signed URLs, and monitor usage — all through natural language prompts.
list_projectsList all image projectscreate_projectCreate a new projectget_projectGet project details & statsupdate_projectUpdate name or domaindelete_projectDelete a projectupload_imageUpload from URL or base64 — handles presigned flow automaticallylist_imagesList images with paginationupdate_imageUpdate alt text and tagsdelete_imageDelete an imagelist_presetsList transform presetscreate_presetCreate a named presetupdate_presetUpdate a presetdelete_presetDelete a presetget_signing_configGet signing configurationupdate_signingEnable/disable URL signinggenerate_signed_urlGenerate HMAC-SHA256 signed URLget_usageDaily requests, bandwidth, transformsbuild_cdn_urlBuild a CDN URL with transforms (no API call)Once connected, just talk to your AI assistant naturally.
MCP is an open standard that lets AI assistants connect to external tools and data sources. It's supported by Claude, GPT, Cursor, Windsurf, and other AI clients. The Spronta MCP Server uses this protocol to give AI agents access to the Spronta Image CDN.
Any MCP-compatible client works, including Claude Code, Claude Desktop, Cursor, Windsurf, and any client that supports the Model Context Protocol. The server uses the standard stdio transport.
The upload_image tool handles the entire flow automatically. Pass an image URL or base64 data, and the server fetches the image, gets a presigned upload URL from Spronta, uploads the file, confirms the upload, and returns the CDN URL with a computed blurhash — all in one tool call.
Yes. The Spronta MCP Server is MIT-licensed and available on GitHub at github.com/spronta/mcp and on npm as @spronta/mcp.
Yes. You need a Spronta account and an API key to authenticate. Sign up at app.spronta.com to get started — the free tier includes 100 images.
Get started in 30 seconds. Free tier available.