MCPH Documentation

What is MCPH?

MCPH is a modern USB key for AI agents—store, share, and auto-expire artifacts in one command. It works with ChatGPT, Claude, and other AI tools that support the Model Context Protocol (MCP).

With MCPH, you can:

• Share AI outputs with anyone via a simple link - no login required to view
• Store multiple content types including markdown, code, JSON, images, and binary files
• Integrate directly with AI tools so you can create and share without manual uploads
• Password-protect sensitive content with a simple security option

MCPH is built on the Model Context Protocol (MCP), which enables AI models to maintain and share context across interactions.

Features

• Hybrid Search

  Combine keyword and semantic search for more relevant results.

Anonymous vs. Logged-in Users

MCPH offers different capabilities depending on whether you are logged in or using the service anonymously.

Anonymous users can only download public crates. Logged-in users can manage their own crates and API keys.

Getting Started

Step 1: Create an Account

Visit mcph.io and sign up for a free account. After logging in, navigate to the API Keys section to generate your personal API key. You'll need this to connect AI tools to your MCPH account.

Step 2: Connect Your AI Tool

MCPH works with any AI tool that supports the Model Context Protocol. Crates are stored for 30 days and all generated download links expires in 24 hours. The simplest way to connect is using the mcp-remote CLI tool:

1. Install mcp-remote (one-time setup):

   npm install -g mcp-remote

   Or use npx to run without installing (shown in examples below).

2. Connect to MCPH with your API key:

   npx mcp-remote https://mcp.mcph.io/mcp --header "Authorization: Bearer YOUR_API_KEY"

   Replace YOUR_API_KEY with the API key you generated in Step 1.

3. Configure your AI tool to use MCPH:

   Different AI tools have different setup methods. Here are a few common ones:

   ChatGPT

   In ChatGPT Teams or Enterprise, use the "Configure Custom Integration" feature and add MCPH as a new action with the endpoint https://mcp.mcph.io/mcp

   Claude

   In Claude Pro or Enterprise, navigate to Settings > File & Storage and add MCPH as a remote MCP server with the URL https://mcp.mcph.io/mcp

   Local AI Tools

   Many local AI tools and editors support configuration files. Add this to your config:

   {
     "mcpServers": {
       "mcph": {
         "command": "npx",
         "args": [
           "mcp-remote", 
           "https://mcp.mcph.io/mcp",
           "--header",
           "Authorization: Bearer YOUR_API_KEY"
         ]
       }
     }
   }

Step 3: Create and Share Content

Once connected, you can ask your AI to create and share content via MCPH. For example:

The AI will create the content and provide you with a shareable link that looks like:https://mcph.io/crate/abc123

Share this link with anyone - they don't need an account to view the content.

Available MCP Tools

Input: { limit?: number, startAfter?: string }
Output: { 
  crates: [{ 
    id: string, 
    title: string, 
    description?: string, 
    category: "markdown" | "code" | "image" | "json" | "binary",
    createdAt: string, 
    mimeType: string,
    size: number,
    tags: string[],
    isPublic: boolean,
    isPasswordProtected: boolean
  }, ...],
  lastCrateId: string,
  hasMore: boolean 
}

Input: { id: string, password?: string }
Output: { 
  content: [{ 
    type: "text" | "image", 
    text?: string,          // For text content
    data?: string,          // For image content (base64)
    mimeType?: string       // For images
  }, ...]
}

Input: { id: string, expiresInSeconds?: number }
Output: { url: string, validForSeconds: number }

Input: { query: string }
Output: { 
  crates: [{ 
    id: string, 
    title: string, 
    description?: string, 
    category: "markdown" | "code" | "image" | "json" | "binary",
    createdAt: string, 
    mimeType: string,
    size: number,
    tags: string[],
    isPublic: boolean,
    isPasswordProtected: boolean
  }, ...]
}

Input: { fileName, contentType, data, ... }
Output (binary): { uploadUrl, fileId, gcsPath, message }
Output (text): { crate, message }

Input: { id: string, password?: string }
Output: { id, isShared, shareUrl, message }

Input: { id: string }
Output: { message, ... }

Input: { id: string }
Output: { message }

Input: { id: string }
Output: { crate, message }

How the MCP Endpoint Works

The MCP endpoint (/api/mcp) is the only supported way to interact with MCPH programmatically. It uses the Model Context Protocol (MCP) over Streamable HTTP for real-time, bidirectional communication.

1. Connect via Streamable HTTP: Use npx mcp-remote https://mcp.mcph.io/mcp or configure your client to use the endpoint.
2. Authentication: Pass your API key as a Bearer token in the Authorization header.
3. Session: On connect, you receive a session ID in the MCP-Session-ID response header. All subsequent requests must include this ID in the MCP-Session-ID header.
4. Calling Tools: Send JSON-RPC requests to the endpoint. Example for crates/list:

   
   {
     "jsonrpc": "2.0",
     "id": 1,
     "method": "tools/call",
     "params": {
       "name": "crates/list",
       "arguments": {}
     }
   }
   

   The response will be streamed using Streamable HTTP, with content delivered as chunks in the HTTP response.

For more details, see the mcp-remote documentation or the MCP protocol reference.

Learn More

• MCP Protocol Overview
• mcp-remote npm package
• mcph.io

Using MCPH with AI Assistants

MCPH tools integrate with AI assistants like Claude and ChatGPT to provide seamless file management through natural conversation. You can manage your files without learning commands or APIs—just speak naturally about what you want to do.

Getting Started with AI Assistants

1. Connect MCPH to Your AI Assistant

   Follow the setup instructions for your specific AI assistant to connect it with your MCPH account using the API key.

2. Use Natural Language Commands

   Interact with your AI assistant using simple language - no need to learn special commands.

   "Show my files", "Search for reports", "Share this document", "Delete old files", etc.
3. Manage Files Seamlessly

   Your AI assistant will handle file management tasks using MCPH, such as storing, sharing, and organizing files.