PlugMyDocs

Welcome to PlugMyDocs

PlugMyDocs is a managed platform designed to enhance AI assistants and AI chatbots by providing seamless access to a central collection of documents.

PlugMyDocs is an MCP Server

MCP is an open, industry-standard protocol that defines how AI applications (clients) request information and capabilities from external services (servers) and vice versa. It was introduced by Anthropic in late 2024 and has since been adopted by major players like OpenAI and Microsoft. Think of MCP as a common language that all AI tools and data sources agree to speak.

Getting Started

1

Sign Up

Create an account through our Auth0-powered authentication system. A free trial of 14 days lets you try all the features.

2

Connect your MCP client

Add our MCP server with just a few lines.

Example: Adding PlugMyDocs to Claude Desktop

Step 2a: Open Claude Desktop Settings

In Claude Desktop, click on the settings icon (gear icon) in the bottom left corner, or use the keyboard shortcut Cmd+, (Mac) or Ctrl+, (Windows/Linux).

Step 2b: Navigate to Developer Tab

In the settings window, click on the “Developer” tab in the left sidebar, then click the “Edit Config” button.

Claude Desktop Developer Settings showing Edit Config button
Step 2c: Add PlugMyDocs Configuration

Add the following configuration to your MCP servers section in the JSON file:

"plugmydocs": {
  "command": "npx",
  "args": [
    "mcp-remote",
    "https://mcp.plugmydocs.com/sse"
  ]
}

Make sure to add a comma after the previous server entry if there are other MCP servers configured.

Note: After saving the configuration, restart Claude Desktop for the changes to take effect. PlugMyDocs will then be available as an MCP server in your Claude conversations.

Other MCP Clients (Cursor, etc.)

You may need to adjust the command and add “-y” to the args line for other MCP clients like Cursor.

Cursor MCP configuration adjustment showing -y argument

Use the following configuration for other MCP clients:

"plugmydocs": {
  "command": "npx",
  "args": [
    "-y",
    "mcp-remote",
    "https://mcp.plugmydocs.com/sse"
  ]
}
3

Upload Documents

Add your documents to the platform for processing. Just tell the LLM you are using to uplaod the document, give it an url link to the document and tell it which Knowledge Base to add it to. (a new KB is created automatically if it doesn’t exist)

4

Start Using PlugMyDocs

Once connected, you can use PlugMyDocs tools directly in your AI conversations to upload documents, query your knowledge base, and manage your document collection. See available tools below

Available MCP Tools

PlugMyDocs exposes a suite of tools that AI assistants can invoke through function calls, enabling dynamic interactions with your document knowledge base. These tools include:

Upload Document

Queue a document for ingestion into a specified knowledge base.

Parameters: url (string), kb_id (string)

List Knowledge Bases

Retrieve a list of available knowledge bases.

Parameters: none

List Documents

List all documents within a specific knowledge base.

Parameters: kb_id (string)

Delete Document

Remove a specific document from a knowledge base.

Parameters: kb_id (string), doc_id (string)

Ask RAG

Submit a query to a knowledge base and receive a contextually relevant answer.

Parameters: query (string), kb_id (string)

Get Job Status

Check the status of a document ingestion job or list recent jobs.

Parameters: job_id (string, optional)

Delete Knowledge Base

Initiate the deletion process for an entire knowledge base.

Parameters: kb_id (string)

Delete All Documents in Knowledge Base

Remove all documents from a knowledge base while keeping the KB itself.

Parameters: kb_id (string)

Confirm Deletion

Confirm and execute a pending deletion operation.

Parameters: pending_deletion_id (string)

These tools are designed to be seamlessly integrated with AI assistants, allowing for automated and intelligent management of your document repositories.

How the MCP works

1. User Authentication

We use Auth0 to ensure secure user authentication. Every time you add the mcp to an AI assistant, you are asked to authorize the AI to access the mcp.

2. Document Processing

Once authenticated, you can upload your documents to PlugMyDocs. Our system processes these documents, preparing them for efficient retrieval and interaction.

3. Retrieval-Augmented Generation (RAG)

PlugMyDocs employs RAG to enhance AI responses:

  • Embedding: Your documents are converted into vector representations using advanced embedding techniques. Vectors embeddings are efficient numerical representations of chunks of media.
  • Storage: These vectors are stored in a vector database, optimized for semantic and traditional keyword search.
  • Query Handling: When your AI assistant receives a query, it retrieves relevant document segments from the vector database to provide accurate and context-rich responses.

This method ensures that your AI assistant can provide answers grounded in your specific content, not the LLM’s training data, improving relevance.

4. Upcoming Feature: RAG-less Search

We’re in the process of integrating a RAG-less search option that leverages large language models (LLMs) with extended context capabilities. This will allow for on-the-fly document processing and citation grounding of documents without relying on pre-processed embeddings and vector database retrieval.

Support and Resources

For assistance and more information:

Feel free to reach out to our support team for any queries or assistance.