Parse Documents with AI Agents Using the Open-Source GroupDocs.Parser Cloud MCP Server

Overview

The GroupDocs.Parser Cloud MCP Server is a lightweight wrapper around the GroupDocs.Parser Cloud API that implements the Model Context Protocol (MCP). By exposing document‑parsing capabilities (text, images, barcodes) and cloud‑storage utilities (list, upload, download, delete) as MCP-compliant tools, the server lets AI agents, assistants, and development tools interact with documents just like they would with any other model‑driven data source. This removes the need for custom SDK calls or proprietary integrations, making it straightforward to embed document extraction into LLM‑driven workflows, autocompletion tools, or code‑assistant environments.

Key benefits include:

• Universal access – Any MCP‑compatible client (VS Code, Cursor, KiloCode, custom agents and others) can call the same endpoint to parse documents stored in GroupDocs Cloud.
• Rich extraction – Retrieve plain text, embedded images, and barcodes from over 50 file formats (PDF, Word, Excel, PowerPoint, emails, archives, etc.).
• Storage operations – List folders, upload new files, download existing ones, and manage cloud storage directly via MCP calls.
• Cross‑platform – Runs on Windows, macOS, and Linux with a single Python‑based service.

Below is a quick navigation to the sections that walk you through the protocol, installation, configuration for popular tools, advanced options, and FAQs.

• What is Model Context Protocol (MCP)?
• Why use GroupDocs.Parser Cloud MCP Server?
• Quick Start
• Using MCP with KiloCode
• Using MCP with Cursor
• Using MCP with VS Code
• Advanced Options
• Conclusion
• See also
• FAQ

What is Model Context Protocol (MCP)?

The Model Context Protocol (MCP) is a standard interface that allows large language models (LLMs) and AI agents to interact with external tools and services in a structured, predictable, and discoverable way.

Instead of embedding business logic directly into prompts, MCP exposes external capabilities (APIs, services, data sources) as typed tools that an AI agent can call when needed.

Key MCP concepts

• Tool-based integration
  Each capability is exposed as a tool with a clear purpose (for example, extract text from a document or list files in storage). AI agents can select and invoke these tools dynamically based on user intent.

• Typed input and output schemas
  MCP tools define their inputs and outputs using JSON schemas. This removes ambiguity, reduces hallucinations, and allows models to reason about which tool to call and how to use the result.

• Explicit separation of reasoning and execution
  The LLM focuses on reasoning and decision-making, while the MCP server performs deterministic operations such as document parsing, file processing, or data retrieval.

• Reusable across environments
  Any MCP-compatible client (AI IDEs, chat applications, autonomous agents, local tools) can connect to the same MCP server without custom glue code.

By implementing MCP, a service becomes AI-native: its functionality can be discovered, understood, and safely invoked by AI agents as part of a larger workflow.

Why use GroupDocs.Parser Cloud MCP Server?

Quick Start

This section shows how to configure and run the GroupDocs.Parser Cloud MCP Server in just a few steps.

1. Clone the repository

git clone https://github.com/groupdocs-parser-cloud/groupdocs-parser-cloud-mcp.git
cd groupdocs-parser-cloud-mcp

2. Configure environment variables

Create a .env file with your GroupDocs Cloud credentials.
You can either create it manually or copy the provided template .env.example.

CLIENT_ID=your-client-id
CLIENT_SECRET=your-client-secret
MCP_PORT=8000

You can obtain your Client ID and Client Secret from the GroupDocs Cloud dashboard:
https://dashboard.groupdocs.cloud/#/applications

3. Run the MCP server

Choose the command matching your operating system.

Linux / macOS

./run_mcp.sh

Windows (PowerShell)

.\run_mcp.ps1

Windows (Command Prompt)

run_mcp.bat

Server endpoint

Once started, the MCP server is available at:

http://localhost:8000/mcp

You can now connect this endpoint to any MCP-compatible host, such as AI agents, IDE copilots, or LLM tools that support the Model Context Protocol.

Using MCP with KiloCode

KiloCode can call any MCP endpoint directly from its chat interface.

Preparation Steps

1. Open KiloCode Settings → MCP Servers.
2. Add a new server entry named groupdocs-parser-mcp-local.
3. Paste the configuration JSON (URL and type).

KiloCode Configuration JSON

{
  "mcpServers": {
    "groupdocs-parser-mcp-local": {
      "type": "streamable-http",
      "url": "http://127.0.0.1:8000/mcp"
    }
  }
}

Example Prompts in KiloCode

Extract all text from sample_invoice.pdf using the GroupDocs.Parser MCP server, then give me a brief summary of the invoice amount

When you send the prompt, KiloCode will:

• Upload sample_invoice.pdf to GroupDocs Cloud.
• Call the MCP parse/text method.
• Return the model‑generated summary.

Extract all images from document.pdf, save them in current folder, subfolder “document_images” and after processing remove the images from GroupDocs.Cloud storage When you send the prompt, KiloCode will:

• Upload document.pdf to GroupDocs Cloud.
• Call the MCP tool to extract images.
• Call the parse/images endpoint via MCP to extract images.
• Call the MCP to download the extracted images and save them into the document_images folder.

Using MCP with Cursor

Cursor’s “Tools & MCP” panel lets you register custom MCP servers.

Setup Steps

1. Open Cursor Settings → Tools & MCP.
2. Click Add Custom MCP.
3. Insert the following JSON snippet into the mcp.json file section.

Cursor mcp.json Configuration

{
  "mcpServers": {
    "groupdocs-parser-mcp-local": {
      "url": "http://127.0.0.1:8000/mcp"
    }
  }
}

Sample Prompt for Cursor

Extract text from Message.msg using the GroupDocs.Parser MCP, then give me a brief summary of the email message.

Cursor will automatically:

• Upload the file in your personal GroupDocs.Cloud storage
• Retrieve text using the GroupDocs.Parser Cloud.
• Include the results in the chat answer.

Using MCP with VS Code

VS Code supports MCP servers. In this section we’ll show you how to set it up and use the GroupDocs MCP server features.

Configuration Steps

1. Create a folder .vscode in your project if it doesn’t exist.
2. Add a file named mcp.json with the server definition.

VS Code mcp.json Example

{
  "servers": {
    "groupdocs-parser-mcp-local": {
      "type": "http",
      "url": "http://127.0.0.1:8000/mcp"
    }
  }
}

3. Reload VS Code (Ctrl+Shift+P → Developer: Reload Window).
4. Now you can open a Chat (Ctrl+Alt+I) and the chat will invoke the MCP tools when requested.

For example, ask a question in chat about your file located in the current folder opened in VSCode:

Parse the Invoice.pdf using the GroupDocs.Parser MCP, give me a brief summary of invoice.

Advanced Options

Test the Server with MCP Inspector

The MCP Inspector is a lightweight UI that lets you explore the server’s schema and try calls interactively.

# Run the inspector (Node.js required)
npx @modelcontextprotocol/inspector

In the browser:

1. Choose “streamable HTTP” as the connection type.
2. Enter your server URL: http://127.0.0.1:8000/mcp.
3. Click Connect and browse available methods (e.g., parser_extract_text, parser_extract_barcodes).

Reinitializing the Virtual Environment

If you modify requirements.txt or encounter dependency errors, reinitialize the environment:

# Linux / macOS
./init_mcp.sh

# Windows PowerShell
.\init_mcp.ps1

# Windows CMD
init_mcp.bat

The script will:

1. Delete the existing .venv.
2. Re‑create a clean virtual environment.
3. Re‑install all packages from requirements.txt.

After the reset, start the server again:

# Linux / macOS
./run.sh

# Windows PowerShell
.\run.ps1

# Windows CMD
run.bat

Conclusion

In this article we covered:

• What MCP is and why it matters for AI‑driven tooling.
• GroupDocs.Parser Cloud MCP Server – a lightweight bridge that adds text, image, and barcode extraction plus full storage management to any MCP‑compatible client.
• Step‑by‑step installation (clone, configure, run).
• How to plug the server into popular environments – KiloCode, Cursor, and VS Code.
• Advanced diagnostics using the MCP Inspector and environment reinitialization.

With the MCP server in place, developers can let LLMs interact with documents as naturally as they do with databases or APIs, eliminating boilerplate SDK code and accelerating AI‑powered document workflows.

See also

• GroupDocs.Parser Cloud API Reference – https://reference.groupdocs.cloud/parser/
• GroupDocs.Parser Cloud Documentation – https://docs.groupdocs.cloud/parser/
• Start a Free Trial – https://purchase.groupdocs.cloud/cloud/trial/
• Purchase Policies and FAQ – https://purchase.groupdocs.cloud/cloud/policies/

Frequently Asked Questions (FAQs)

• Q: Why is the MCP server open-source and run locally? Why isn’t there a publicly hosted MCP server?
  A: The Model Context Protocol (MCP) is still a very new standard and current LLMs and AI assistants typically have limited or unreliable native support for binary file streams (PDFs, images, archives). Document parsing, OCR, image extraction, and barcode recognition are areas where specialized APIs like GroupDocs.Parser Cloud excel. The local MCP server bridges this gap in a reliable and standardized way.

• Q: Do I need to install any additional software to use the MCP server?
  A: No. The server runs on any OS that supports Python 3.10+ and requires only the packages listed in requirements.txt.

• Q: Which document formats are supported?
  A: Over 50 formats, including PDF, DOCX, XLSX, PPTX, email (.eml, .msg), archives (ZIP, RAR), and common image types (PNG, JPG, TIFF).

• Q: Can I extract barcodes from scanned PDFs?
  A: Yes. The MCP server supports the parse/barcodes endpoint that detects 1D and 2D barcodes in raster images and PDFs.

• Q: How do I list files in a specific GroupDocs Cloud folder?
  A: The MCP server supports GroupDocs.Cloud storage endpoints (storage/list, storage/upload, storage/download, storage/delete) and the storage operations are used in chat sessions automatically or by your request.

• Q: What if I change the MCP port after the server is running?
  A: Update the MCP_PORT value in .env and restart the server (run.sh / run.ps1).

• Q: Do I get a free trial?
  A: Yes, 150 free API calls per month are available.