dragdropdo
High-Performance and Unlimited File Conversion MCP
MCP clients (Claude Code, Cursor, custom)
Connect AI assistants and applications to DragDropDo using the Model Context Protocol (MCP) over HTTPS.
DragDropDo MCP server
The DragDropDo MCP server exposes the same capabilities as the Business API—upload, convert, compress, merge, zip, PDF password tools, and status polling—as MCP tools. Clients connect over Streamable HTTP (HTTPS).
Prerequisites
- A D3 API key from the dashboard (sign in → Account → generate API key). Same key as for
Authorization: Beareron/api/v1. - MCP URL:
https://mcp.dragdropdo.com/mcp
Send the API key on every MCP request in the Authorization header:
Authorization: Bearer <YOUR_D3_API_KEY>
The server validates this key the same way as REST requests (see Authentication).
Claude Code
Claude Code supports remote MCP servers over HTTP. Prefer --transport http (SSE is deprecated).
CLI (recommended)
Add the server and pass your API key on the Authorization header:
claude mcp add --transport http d3 https://mcp.dragdropdo.com/mcp \
--header "Authorization: Bearer YOUR_D3_API_KEY"
Useful commands: claude mcp list, claude mcp get d3, claude mcp remove d3.
Scopes: add --scope project to store the entry in the project’s .mcp.json, or --scope user for your user config. See Connect Claude Code to tools via MCP.
JSON configuration
Claude Code merges MCP entries from ~/.claude.json and, for project scope, .mcp.json. An HTTP server entry looks like this (structure may vary slightly by Claude Code version—use claude mcp get <name> after adding to see the exact shape):
{
"mcpServers": {
"d3": {
"type": "http",
"url": "https://mcp.dragdropdo.com/mcp",
"headers": {
"Authorization": "Bearer YOUR_D3_API_KEY"
}
}
}
}
You can use environment substitution in values where your tooling supports it (for example Bearer ${D3_API_KEY}), so secrets are not committed to git.
Cursor
In Cursor, MCP servers are defined in the MCP config file (often user: ~/.cursor/mcp.json, or project .cursor/mcp.json depending on your setup).
Add an entry with url and headers:
{
"mcpServers": {
"d3": {
"url": "https://mcp.dragdropdo.com/mcp",
"headers": {
"Authorization": "Bearer YOUR_D3_API_KEY"
}
}
}
}
Restart Cursor or reload MCP servers so the new server is picked up. The assistant can then call DragDropDo tools (for example initiate_upload, convert_file, poll_status) when appropriate.
Custom MCP clients
Use an official Model Context Protocol client with Streamable HTTP to https://mcp.dragdropdo.com/mcp, authenticated with Authorization: Bearer and your API key.
The examples below initialize the session and list tools. Match imports and APIs to your SDK version.
```ts NODE theme={null} import { Client } from "@modelcontextprotocol/sdk/client/index.js"; import { StreamableHTTPClientTransport } from "@modelcontextprotocol/sdk/client/streamableHttp.js";const url = new URL("https://mcp.dragdropdo.com/mcp"); const transport = new StreamableHTTPClientTransport(url, { authProvider: { token: async () => process.env.D3_API_KEY!, }, });
const client = new Client({ name: "my-app", version: "1.0.0" }); await client.connect(transport); await client.listTools(); // await client.callTool({ name: "supported_operation", arguments: { ext: "pdf" } });
```python PYTHON theme={null}
import os
import httpx
from mcp import ClientSession
from mcp.client.streamable_http import streamable_http_client
URL = "https://mcp.dragdropdo.com/mcp"
api_key = os.environ["D3_API_KEY"]
headers = {"Authorization": f"Bearer {api_key}"}
async def main():
async with httpx.AsyncClient(headers=headers) as http_client:
async with streamable_http_client(URL, http_client=http_client) as (read, write):
async with ClientSession(read, write) as session:
await session.initialize()
await session.list_tools()
# await session.call_tool("supported_operation", {"ext": "pdf"})
# asyncio.run(main())
<?php
declare(strict_types=1);
use Mcp\Client;
use Mcp\Client\Transport\HttpTransport;
$apiKey = getenv('D3_API_KEY') ?: '';
$client = Client::builder()
->setClientInfo('My App', '1.0.0')
->build();
$transport = new HttpTransport(
endpoint: 'https://mcp.dragdropdo.com/mcp',
headers: ['Authorization' => 'Bearer ' . $apiKey],
);
$client->connect($transport);
$client->listTools();
// $client->callTool(name: 'supported_operation', arguments: ['ext' => 'pdf']);
$client->disconnect();
# Gemfile: gem "mcp" / gem "faraday", ">= 2.0"
require "mcp"
http_transport = MCP::Client::HTTP.new(
url: "https://mcp.dragdropdo.com/mcp",
headers: {
"Authorization" => "Bearer #{ENV.fetch('D3_API_KEY')}",
},
)
client = MCP::Client.new(transport: http_transport)
client.tools
# client.call_tool(tool: client.tools.first, arguments: { "ext" => "pdf" })
package main
import (
"context"
"os"
"github.com/mark3labs/mcp-go/client"
"github.com/mark3labs/mcp-go/client/transport"
"github.com/mark3labs/mcp-go/mcp"
)
func main() {
ctx := context.Background()
key := os.Getenv("D3_API_KEY")
httpTransport, err := transport.NewStreamableHTTP(
"https://mcp.dragdropdo.com/mcp",
transport.WithHTTPHeaders(map[string]string{
"Authorization": "Bearer " + key,
}),
)
if err != nil {
panic(err)
}
c := client.NewClient(httpTransport)
defer c.Close()
if err := c.Initialize(ctx); err != nil {
panic(err)
}
if _, err := c.ListTools(ctx, mcp.ListToolsRequest{}); err != nil {
panic(err)
}
// c.CallTool(ctx, mcp.CallToolRequest{ ... })
}
import io.modelcontextprotocol.sdk.client.McpClient;
import io.modelcontextprotocol.sdk.client.McpSyncClient;
import io.modelcontextprotocol.sdk.client.transport.HttpClientStreamableHttpTransport;
import io.modelcontextprotocol.sdk.spec.McpTransport;
// io.modelcontextprotocol.sdk:mcp — use your BOM or version from https://java.sdk.modelcontextprotocol.io/
String apiKey = System.getenv("D3_API_KEY");
McpTransport transport = HttpClientStreamableHttpTransport.builder("https://mcp.dragdropdo.com")
.endpoint("/mcp")
.requestCustomizer(req -> req.header("Authorization", "Bearer " + apiKey))
.build();
McpSyncClient client = McpClient.sync(transport).build();
client.initialize();
client.listTools();
// client.callTool(new CallToolRequest("supported_operation", Map.of("ext", "pdf")));
client.closeGracefully();
SDK references: TypeScript, Python, PHP mcp/sdk, Ruby mcp gem, Go mark3labs/mcp-go, Java io.modelcontextprotocol.sdk:mcp.
Available tools (summary)
Typical tool names include: initiate_upload, complete_upload, supported_operation, convert_file, compress_file, merge_files, zip_files, lock_pdf, unlock_pdf, reset_pdf_password, get_status, poll_status, get_download_link. The server also sends instructions describing the upload and operation workflow. Use list_tools from your client after connecting to see the canonical list and schemas for your session.
Related documentation
- Business API overview — REST surface parallel to MCP tools
- Authentication — API key usage and security
- Quickstart — upload and operation flow with REST examples
相关服务器
Geo Location Demo
Retrieves user geolocation information using EdgeOne Pages Functions and exposes it via an MCP server.
Salesforce TypeScript Connector
Interact with Salesforce data using SOQL queries, SOSL searches, and CRUD operations via a TypeScript MCP server.
Edgee MCP Server
Manage organizations, projects, components, and users via the Edgee API.
MCP Server To Markdown
Converts various file formats to Markdown using Cloudflare AI.
NFTGo MCP
Access the NFTGo Developer API for comprehensive NFT data and analytics. Requires an NFTGo API key.
PayPal by CData
A read-only MCP server for querying live PayPal data, powered by the CData JDBC Driver.
Zen MCP Server
Access multiple AI models like Claude, Gemini, and OpenAI through a single server using your own API keys.
Dokploy
An AI-powered interface for managing the Dokploy infrastructure platform.
Cloudflare MCP Server
A self-hostable MCP server for deployment on Cloudflare Workers using Durable Objects.
ProbeOps
Network diagnostics from 6 global regions — SSL, DNS, ping, whois, traceroute, port scan, latency. Works without an API key.