FortiPhrase - Password Strength Analyzer & Passphrase Generator

FortiPhrase is a local, offline tool designed to help users assess the strength of their passwords, generate secure passphrases, and perform bulk password analysis. By using Generative AI (GenAI) and Machine Learning (ML) techniques, the tool goes beyond basic password metrics (e.g., length, special characters) to evaluate overall password strength and predict vulnerabilities to common attack methods. All analysis is done locally without storing or transmitting any data.

Features

1. Password Strength Analyzer & Suggester

• Evaluates Strength: Analyzes password strength using 15 features (e.g., length, symbols, character variety, etc.) and a decision tree regressor.
• Strength Score: Provides a strength score ranging from 0 to 1.
• Time-to-Crack Estimates: Estimates the time required to crack a password using dictionary, rainbow table, or brute force methods.
• AI Suggestions: Local AI (using Ollama and LLaMA3 models) suggests improvements to make passwords stronger.

2. Passphrase Generation

• Generates Secure Passphrases: Creates strong passphrases based on personal keywords.
• Personalized Suggestions: Tailored to avoid password reuse while remaining memorable and secure.

3. Bulk Password Analysis

• Multiple Password Analysis: Analyze several passwords simultaneously for strength and vulnerability.
• Strength and Crack-Time Estimates: Provides the strength score and time-to-crack estimates for each password.
• Strong Suggestions: Flags weak passwords and provides stronger alternatives.

4. Bulk Password Analyzer with PII Matching

• Detects Sensitive Data: Scans passwords for Personally Identifiable Information (PII).
• CSV File Support: Processes passwords from CSV files and flags any that contain sensitive data.

ML Model used

Our model is a Decision Tree Regressor that predicts password strength based on 15 password characteristics namely:

Feature	Description
len	Total number of characters in the password.
alphaUC	Count of uppercase letters.
alphaLC	Count of lowercase letters.
number	Count of digits.
symbol	Count of special characters (!@#$%^&*).
midChar	Count of digits or special characters in the middle of the password.
repChar	Number of repeating characters (length minus unique characters).
uniqueChar	Number of distinct characters.
consecAlphaUC	Number of consecutive uppercase letters.
consecAlphaLC	Number of consecutive lowercase letters.
consecNumber	Number of consecutive digits.
consecSymbol	Number of consecutive special characters.
seqAlpha	Number of sequential letter patterns (e.g., "abc" or "cba").
seqNumber	Number of sequential digit patterns (e.g., "123" or "321").
seqKeyboard	Number of sequential keyboard patterns (e.g., "qwe" or "ewq").

Setup

Clone the Repository

git clone https://github.com/Saur-Deshmukh/HackOHire-HashHackers.git

Ollama

1. Install Ollama from here
2. Pull llama3 onto your system:

   ollama pull llama3

3. Run Ollama:

   ollama run llama3

Frontend (Next.js)

1. Change Directory:

   cd Frontend

2. Install dependencies:

   npm install --legacy-peer-deps

3. Start the development server:

   npm run dev

Backend (Flask)

1. Setup Virtual Environment:

   python -m venv myenv
   myenv\Scripts\activate

2. Install dependencies:

   pip install -r requirements.txt

3. Change Directory:

   cd Backend

4. Start the servers:

   python analyze.py
   python bulk_pii.py
   python bulk.py
   python insights.py
   python passphrase.py
   python time_to_crack.py

Services and Ports

Service	File	Localhost Port
Bulk Operations	bulk.py	http://localhost:5000
PassPhrase Generator	passphrase.py	http://localhost:5001
Password Strength Analyzer	analyser.py	http://localhost:5002
Security Insights and Reports	insights.py	http://localhost:5003
Time to Crack Estimator	time_to_crack.py	http://localhost:5004
Bulk PII Handler	bulk_pii.py	http://localhost:5006

Privacy

• Offline Processing: All password analysis is done locally on your machine. No data is stored or transmitted.
• Local AI Models: Uses Ollama and LLaMA3 AI models for password strength analysis and suggestions, ensuring complete privacy.

Technologies Used

• Frontend: Next.js
• Backend: Python, Flask
• AI Models: Ollama, LLaMA3
• Machine Learning Techniques: Decision tree regressor
• Additional Libraries: CSV processing, PII detection

Acknowledgments

The FortiPhrase tool uses the Ollama and LLaMA3 AI models for offline password strength analysis and suggestions.