ThisAkshat/PHANTOM-AI-Cyber-Threat-Analyzer
PHANTOM is an AI-powered phishing detection system that analyzes emails, messages, and URLs using NLP, machine learning, and cyber-threat intelligence to identify scams in real time.
What's novel
PHANTOM is an AI-powered phishing detection system that analyzes emails, messages, and URLs using NLP, machine learning, and cyber-threat intelligence to identify scams in real time.
Code Analysis
15 files read · 4 roundsA Streamlit app that fine-tunes BERT on 15 rows of phishing text, then combines the ML prediction with hardcoded keyword matching, WHOIS lookups, SSL checks, and brand-spoof heuristics to produce a phishing/safe verdict.
Strengths
The multi-layer decision pipeline concept (ML + domain age + SSL + brand spoof + ASN) mirrors how real security products think, and the brand_intel module with official domain mapping and suspicious TLD detection is the most substantive piece of logic.
Weaknesses
The training dataset is 15 rows (far too small for BERT fine-tuning), the AbuseIPDB API key is empty so IP reputation always returns 0, the OpenPhish feed is fetched but never used in the decision pipeline, there's a duplicate hard-block loop in predict.py, bare except clauses swallow all errors, and the 'test' is a single print statement with no assertions.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
13
Contributors
1
Files
21
Active weeks
3
Repository
Language
Python
Stars
2
Forks
0
License
—