What is Agentic Ai Tool Suite

The agentic-ai-tool-suite is a collection of Model Context Protocol (MCP) servers designed to enhance the capabilities of AI agents by providing tools for media search, information retrieval, PDF generation, and PowerPoint presentation creation.

Use cases

Use cases include generating educational materials, conducting media searches for content analysis, creating professional presentations, and automating document generation from web content.

How to use

To use the agentic-ai-tool-suite, you need to clone the repository, set up each server individually, and configure them within your MCP client (e.g., Cline or Claude Desktop App). Each server runs as a separate process.

Key features

Key features include media tools for searching and understanding images and videos, information retrieval tools for web searches, a PDF creator for generating documents from HTML, and a presentation creator for assembling PowerPoint presentations.

Where to use

The agentic-ai-tool-suite can be used in various fields such as education for creating presentations, media analysis for understanding content, and research for retrieving information from the web.

Unified MCP Suite

This repository contains a collection of Model Context Protocol (MCP) servers, packaged together for convenience. Each server provides distinct functionalities related to media handling, information retrieval, and document creation.

Important: These servers are designed to run as separate processes. You need to set up and configure each server individually within your MCP client (e.g., Cline or the Claude Desktop App).

Included Servers

• Media Tools Server (media-tools-server): Provides tools for searching images (Unsplash) and videos (YouTube), downloading images, and understanding image/video content (Google Gemini, YouTube Transcripts). (Node.js/TypeScript)
• Information Retrieval Server (information-retrieval-server): Provides tools for performing web searches (Google Custom Search) and crawling web pages. (Node.js/TypeScript)
• PDF Creator Server (pdf-creator-server): Provides a tool to generate PDF documents from HTML content using Playwright and Pillow. (Python)
• Presentation Creator Server (presentation-creator-server): Provides tools to assemble PowerPoint presentations from HTML slide content and generate PDFs from HTML. (Python)

Prerequisites

• Node.js and npm: Required for media-tools-server and information-retrieval-server. Download from https://nodejs.org/
• Python and pip: Required for pdf-creator-server and presentation-creator-server. Download from https://www.python.org/
• Git: For cloning this repository.
• MCP Client: An application capable of running MCP servers (e.g., Cline, Claude Desktop App).

Setup Instructions

Follow the steps below for each server you wish to use.

1. Clone the Repository

git clone <repository-url> # Replace <repository-url> with the actual URL
cd unified-mcp-suite

2. Media Tools Server (media-tools-server)

(Node.js/TypeScript)

a. Navigate to Directory:

cd media-tools-server

b. Install Dependencies:

npm install

c. Build the Server:

npm run build

(This compiles the TypeScript code into JavaScript in the build directory.)

d. Obtain API Keys:

• Unsplash API Key: Required for image search. Create an account and register an application at https://unsplash.com/developers.
• Google API Key: Required for video search/understanding and image understanding.
  • Enable the “YouTube Data API v3” and “Vertex AI API” (for Gemini) in your Google Cloud Console project: https://console.cloud.google.com/
  • Create an API key under “APIs & Services” -> “Credentials”. Restrict the key if necessary.

e. Configure MCP Client:

Add the following configuration to your MCP client’s settings file (e.g., cline_mcp_settings.json or claude_desktop_config.json). Replace placeholders with your actual API keys and the correct absolute path to the built server file.

f. Navigate Back:

cd ..

3. Information Retrieval Server (information-retrieval-server)

(Node.js/TypeScript)

a. Navigate to Directory:

cd information-retrieval-server

b. Install Dependencies:

npm install

c. Build the Server:

npm run build

d. Obtain API Keys:

• Google Custom Search API:
  • Enable the “Custom Search API” in your Google Cloud Console project: https://console.cloud.google.com/
  • Create an API key under “APIs & Services” -> “Credentials”.
  • Create a Custom Search Engine and get its ID: https://programmablesearchengine.google.com/

e. Configure MCP Client:

Add the following configuration to your MCP client’s settings file. Replace placeholders with your actual API key, Search Engine ID, and the correct absolute path.

f. Navigate Back:

cd ..

4. PDF Creator Server (pdf-creator-server)

(Python)

a. Navigate to Directory:

cd pdf-creator-server

b. Create and Activate Virtual Environment:

# Create environment
python -m venv .venv

# Activate environment
# Windows (Command Prompt/PowerShell)
.\.venv\Scripts\activate
# macOS/Linux (bash/zsh)
# source .venv/bin/activate

c. Install Dependencies:

pip install -r requirements.txt

d. Install Playwright Browsers: (Required for PDF generation)

playwright install

e. Configure MCP Client:

Add the following configuration to your MCP client’s settings file. Replace the placeholder with the correct absolute path to the Python executable within the virtual environment and the server script.

f. Deactivate Virtual Environment (Optional):

deactivate

g. Navigate Back:

cd ..

5. Presentation Creator Server (presentation-creator-server)

(Python)

a. Navigate to Directory:

cd presentation-creator-server

b. Create and Activate Virtual Environment:

# Create environment
python -m venv .venv

# Activate environment
# Windows (Command Prompt/PowerShell)
.\.venv\Scripts\activate
# macOS/Linux (bash/zsh)
# source .venv/bin/activate

c. Install Dependencies:

pip install -r requirements.txt

d. Install Playwright Browsers: (Required for screenshotting HTML slides)

playwright install

e. Configure MCP Client:

Add the following configuration to your MCP client’s settings file. Replace the placeholder with the correct absolute path to the Python executable within the virtual environment and the server script.

f. Deactivate Virtual Environment (Optional):

deactivate

g. Navigate Back:

cd ..

Running the Servers

Once configured in your MCP client, the servers should start automatically when the client launches. You can then use the tools provided by each server through your MCP client interface.

Example Configuration File

An example configuration file, example_cline_mcp_settings.json, is included in the root of this repository. You can use this as a template for configuring the servers in your MCP client (like Cline).

To use the example:

1. Locate your MCP client’s actual settings file (e.g., c:\Users\YourUser\AppData\Roaming\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json for Cline in VS Code on Windows).
2. Copy the server configurations from example_cline_mcp_settings.json into your actual settings file, merging them with any existing server configurations you might have.
3. Crucially, update all placeholder paths (e.g., C:/absolute/path/to/...) to reflect the actual absolute path where you cloned the unified-mcp-suite repository on your system.
4. Replace all placeholder API keys (e.g., YOUR_GOOGLE_API_KEY_HERE) with your own keys obtained during the setup steps.
5. Ensure the command path for the Python servers correctly points to the python.exe (Windows) or python (macOS/Linux) executable inside the respective .venv directory you created.