About Trusul
Trusul helps users analyze news, claims, scams, and suspicious online content before they trust or share it.
Product Vision
Trusul is being built as a practical trust layer for the modern internet. The goal is simple: help people pause, inspect, and understand suspicious content before forwarding it, reposting it, or acting on it.
The platform combines language analysis, claim extraction, source checks, risk scoring, and clear explanations so users can see why content looks credible, questionable, or high risk.
How Trusul Works
Paste article text or a public URL. Future versions can expand this to screenshots, social posts, and scam messages.
The system checks sensational wording, emotional manipulation, clickbait phrases, ALL-CAPS patterns, and punctuation pressure.
Trusul detects topic, reading level, sentiment patterns, extracted claims, and source coverage indicators.
Users receive a credibility score, risk verdict, explanation, evidence breakdown, and exportable report for later review.
Current Product Capabilities
Trust Scoring
Turns multiple risk signals into a clear 0-100 credibility score, where higher means more credible.
Suspicious Signals
Highlights manipulation patterns commonly found in misleading online content.
Dashboard
Shows user history, trends, recent checks, and topic-level analysis activity.
Compare Mode
Compares two articles or claims side by side for faster credibility review.
Technology Stack
HTML5, CSS3, vanilla JavaScript, and Chart.js.
Python FastAPI, SQLAlchemy, SQLite, JWT authentication, and bcrypt.
NLP heuristics, NLTK, Trafilatura extraction, and optional external verification APIs.
Cloud database, subscription plans, usage limits, API access, and public deployment.