RAG

Provides a Python-based MCP server to manage organizational content via a content service API, including listing, searching, uploading, and deleting content.

python

0

GitHub Stars

python

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "yogeshkulkarni553-rag-mcp-py": {
      "command": "uv",
      "args": [
        "run",
        "mcp_server.py"
      ],
      "env": {
        "USER_ID": "YOUR_USER_ID",
        "CONTENT_SERVICE_URL": "http://localhost:8080"
      }
    }
  }
}

You run a Python-based MCP server that implements the Retrieval-Augmented Generation (RAG) Model Context Protocol (MCP). It provides tools to manage organizational content through a content service API, enabling you to list, search, upload, and delete content as part of a scalable content-management workflow.

How to use

To use this MCP server, you connect an MCP client to the local server and issue tool calls like listing content, performing semantic searches, or uploading content files and URLs about your organization. The server exposes a set of predefined tools you can invoke from your client, and it handles the actual communication with the content service API behind the scenes.

How to install

Prerequisites include a Python environment and the MCP tooling flow used to run servers with an integrated package manager.

# Install uv (modern Python package manager) if you do not have it yet
curl -LsSf https://astral.sh/uv/install.sh | sh
# or
pip install uv

Configuration

Configure the runtime by providing the identifiers and the content service location the MCP server should use.

USER_ID=your_user_id_here
CONTENT_SERVICE_URL=http://localhost:8080

Usage notes

You can start the MCP server using either the uv command directly or a Python invocation, depending on your preferred workflow. The server will listen for MCP protocol messages on stdin/stdout.

Two common ways to start the server are shown below. Choose the one you prefer and ensure all required environment variables are set.

Tools and how they work

listContentNames — List all content names from the organization’s database, optionally filtered by name. You can refine the results by providing a name filter to limit the returned names.
searchOrganizationContents — Perform semantic search across the organization’s content database to locate relevant items based on a query.
deleteOrganizationContent — Remove a specific piece of content from the organization’s database using its content identifier.
uploadContentFileAboutOrganization — Upload a content file about the organization. Provide the file content and a file name, and optionally specify the role of the user.
uploadContentUrlAboutOrganization — Upload a content URL about the organization. Provide the URL and optionally specify the user’s role.

Environment and startup options

Environment variables you set control which user and content service instance the server uses.

Testing

You can run tests to verify behavior and ensure the MCP server interacts correctly with the content service. Use the provided test scripts or the uv-based test runner in your workflow.

Development notes

Development tooling includes code formatting, linting, and type checking. You can format with Black, lint with Ruff, and type check with Mypy to maintain code quality as you develop and extend the server.

Available tools

listContentNames

List all content names from the organization's database optionally filtered by name

searchOrganizationContents

Search through the organization's content database using semantic search

deleteOrganizationContent

Delete specific content from the organization's database

uploadContentFileAboutOrganization

Upload content file about the organization

uploadContentUrlAboutOrganization

Upload content URL about the organization