🔬 AI Research CLI Agent

A powerful command-line research assistant built with Ollama Cloud API.
It searches the web, summarizes results using AI, translates the content, and exports it into multiple file formats.

---

🚀 Features

• 🔍 Web search via DuckDuckGo (DDGS)
• 🧩 AI-powered summarization using Ollama Cloud models
• 🌐 Multi-language translation
• 📄 Export to: TXT, MD, DOCX, PDF (RTL supported)
• 🔐 Secure API key validation
• 🗂 Automatic logging system
• 🎯 Interactive CLI menus

---

---

📦 Installation

Clone the repository:

git clone https://github.com/Mahdi-Haqiqat/AI-Research-Agent.git  
cd AI-Research-Agent

Install dependencies:

pip install -r requirements.txt  

Or install manually:

pip install ollama ddgs questionary python-dotenv httpx python-docx reportlab arabic-reshaper python-bidi  

Note: If you want to install on mobile read Mobile Installation

---

⚙️ Requirements

• 🐍 Python 3.9+
• 🌐 Internet connection

📚 Python dependencies:

ollama
ddgs
questionary
python-dotenv
httpx
python-docx
reportlab
arabic-reshaper
python-bidi  

🔑 API Key

Get your API key from: Ollama

---

▶️ Usage

Run the program:

python main.py  

Then:

1. Enter an API
2. Enter a topic to research
3. Choose target language
4. Choose output format

Or:

Interactive mode (prompts for topic, language, and format)

usage: ai-research [-h] [-l LANGUAGE] [-f FORMAT] [-v] [-c OUTPUT_PATH] [topic ...]

🔬 AI Research Agent - Automated research using AI

positional arguments:

topic Topic to research

options:

-h, --help show this help message and exit

-l, --language LANGUAGE Output language (example: persian, english)

-f, --format FORMAT Output format (txt, md, docx, pdf)

-v, --version show program's version number and exit

-c, --change OUTPUT_PATH Change output directory

e.g,:

python main.py Github -l english -f docx

The tool will:

• Search the web
• Summarize the results
• Translate (if needed)
• Generate the output file
• Log execution details

---

📝 Logging System

All executions are logged inside the /logs/ directory.

Each topic generates its own log file containing:

• Date & Time
• Topic
• Language
• Output format
• Execution status
• Error details (with full traceback if failure occurs)

---

🌍 Supported Languages

English  
Persian  
French  
German  
Arabic  
Spanish  
Italian  
Turkish  

PDF output supports RTL languages (Persian, Arabic, Turkish).

---

⚠️ Important Notes

• PDF generation requires DejaVuSans.ttf in the project root directory.
• Internet connection is required.
• Ollama Cloud models must be accessible in your account.
• Ensure model names in .env are correct.

---

🧠 Project Architecture

search_agent() → Web search
summarize_agent() → AI summarization
translator_agent() → AI translation
writer_agent() → File export
log_agent() → Execution logging

---

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.