Intelligent Content Organizer MCP Agent

Intelligent Content Organizer MCP Agent

Indexes PDF, text, Word and image files, then searches them by meaning, summarizes, tags, categorizes and answers questions. An MCP server with a Gradio page, for anyone with many documents.

Made by Agents-MCP-Hackathon on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space is an MCP server with a Gradio web page for managing a pile of documents. You upload PDF, TXT or DOCX files, or images, and it indexes them for search by meaning. It can also summarize content, generate tags, sort items into categories, answer questions about your files and report statistics on your content.

Who it suits: It suits people who keep many documents and want to search them in plain language, and developers who want an MCP server that offers document tools to an AI assistant.

Industry | Who uses it | How they use it | Example
Research | Analysts | Index papers and reports, then ask questions about them | An analyst asks what a set of reports says about one topic
Business | Operations teams | Sort and tag incoming files into categories | A team categorizes a batch of uploaded documents
Software | Developers | Offer document tools to an MCP client | A developer connects the server to an assistant
Education | Students | Summarize study notes in a chosen style | A student asks for a bullet point summary of lecture notes

You can skip it if you do not want to supply API keys for the model providers it uses.

Where to use it: You use it in these places.

• A web browser, on the Space's page on Hugging Face
• The Gradio page of a local copy on Windows, macOS or Linux
• An MCP client that connects to the server the app provides

Use case | Who | What they get back
Search your files | Anyone with many documents | Matching passages found by meaning, each with a score
Summarize | Readers and analysts | A summary in a concise, detailed, bullet point or executive style
Ask questions | Researchers | Answers drawn from your own documents with retrieval
Organize | Teams | Tags and categories for your content

How to set it up: Online use needs only a browser. A local copy needs Python and provider API keys.

• Python 3.9 or newer, for a local copy
• API keys for OpenAI and Mistral AI, plus an Anthropic API key
• Python packages from requirements.txt

1. Open the Space's page on Hugging Face to use it online, with nothing to install.
2. For a local copy, install the Hugging Face Hub package.
3. Download the Space's files into a folder and move into it.
4. Install Gradio 5.32.0 and the packages in requirements.txt.
5. Make your provider API keys available to the app. The source does not name the settings, so check config.py in the folder.
6. Start the app.

pip install -U huggingface_hub
hf download Agents-MCP-Hackathon/Intelligent_Content_Organizer --repo-type space --local-dir Intelligent_Content_Organizer
cd Intelligent_Content_Organizer
pip install gradio==5.32.0
pip install -r requirements.txt
python app.py

How to use it: Start the app and open the Gradio page it serves. Upload a document, then search it, summarize it, tag it or ask questions about it. An MCP client can call the same tools.

python app.py

Tool | What it does
ingest_document | Processes and indexes a document so it can be searched. Takes a file path and an optional file type.
semantic_search | Searches indexed content with a plain language query. Takes the query, an optional result count with a default of 5, and optional filters.
summarize_content | Summarizes text or an indexed document in a style you pick. The model setting accepts auto or a model family prefix.

Example: You upload a PDF, then run a semantic search for a topic in plain words. The tool returns a list of results, each with its content and a score, plus the total number found.

For newcomers and experts: An MCP server is a program an AI assistant can call as a set of tools. Here the tools handle documents. Try this first: open the Space's page, upload one short text file and ask for a summary.

Under the hood, documents are split into chunks, embedded with Sentence Transformers and searched with FAISS. Mistral OCR reads text out of images. For generation, the auto setting tries OpenAI first, then Mistral, then Anthropic, and a prefix such as gpt-, mistral- or claude- picks a family. The Space's record also names openai/gpt-oss-120b and all-MiniLM-L6-v2.

Upload a document
Text is extracted, with OCR for images
Chunks are embedded and indexed with FAISS
You search, summarize, tag or ask
An LLM provider writes the answer

Good to know: Most features depend on outside model providers, so plan for keys before you start.

• License: mit.
• Cost to run: API keys for OpenAI, Mistral AI and Anthropic, which the prerequisites name. The Space itself uses cpu-basic hardware at Hugging Face.
• Data it reaches: the files you upload, which are PDF, TXT, DOCX and images, and the model providers it calls.
• Not for: use without provider keys, since generation relies on those services.

On Hugging Face, 13 November 2026

• Likes: 16
• Built with: gradio
• License: mit

Links

• Space: https://huggingface.co/spaces/Agents-MCP-Hackathon/Intelligent_Content_Organizer
• Files: https://huggingface.co/spaces/Agents-MCP-Hackathon/Intelligent_Content_Organizer/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.