Skip to content

sigfault-byte/LegitURL

Repository files navigation

LegitURL

A security nutrition label for links — a fully on-device URL scanner that performs over 100 deterministic checks in ≈2 seconds.

Release iOS App Store License

Quick 8-second demo: paste link → score → security findings

Overview

LegitURL is a lightweight mobile app that analyzes the trustworthiness of any URL using a transparent, heuristic-driven approach. All scans are conducted locally and completed in ~2 seconds. No cloud analysis, no data leaks. just fast, explainable results.

Key features:

  • Instant risk scoring – assigns 🟩/🟧/🟥 based on 100+ deterministic checks
  • Security-focused – detects silent redirects, CSP misconfigurations, suspicious TLS certificates, and tracking behavior
  • Explainable results – every finding is traceable to a rule; no black-box logic
  • Privacy-first design – a single HTTPS request, no third-party traffic, zero analytics
  • Exportable reports – generate PDFs or LLM-ready JSON for external review

Media Coverage

Cyberdefense Magazine (July 2025)

"LegitURL offers a unique approach to link analysis — blending pedagogy and precision in a tool designed for everyone."

Excerpt from the article

[...]
But encryption is not authentication.
Rendering is not endorsement.
Even seemingly benign links can conceal redirect chains, cloaked infrastructure, or misconfigured policies — all while wearing the lock like a badge.
I often tell non-technical users to imagine a website as a shop, and their browser as a guide or bodyguard.
That guide will help them get inside, translate unknown languages, and smooth over bumps in the experience.
But how many of us would willingly enter a shop with crumbling walls, broken stairs, sticky notes slapped on our chest, and strangers watching our every move, while the bodyguard just smiles and quietly patches the walls?
[...]

Read the full article page 258.


ZATAZ Cybersecurity News

LegitURL was also featured on ZATAZ, in an article by Damien Bancal, highlighting the tool's unique approach to phishing and scam link detection.


Scoring System

Score Description
🟥 High risk Multiple critical signals: expired/mismatched certs, missing CSP, scam patterns, cloaking, etc.
🟧 Moderate risk Mixed or partial protection. Often seen with major brands but warrants caution.
🟩 Low risk Clean redirect flow, strong TLS, proper headers, no tracking or obfuscation detected.

Getting Started

End-users Download via the App Store
Developers Open LegitURL.xcodeproj in Xcode and build directly.

Screenshots

Signals & Logs Signals and logs view showing coloured findings
Inline script findings Inline script detail with extracted snippet of risky functions
More screenshots
Cookie view Cookie detail with bit-flag severity pyramid
CSP directives Content-Security-Policy directive list
HTML report export Preview of generated HTML security report
LLM JSON export Screen showing compact JSON export for LLMs

How it works

  1. Offline static parsing
    Detects homograph attacks, encoded words, scam phrases, entropy anomalies, and more.

  2. Sandboxed HTTPS fetch
    Retrieves headers, HTML body, TLS certificate, cookies, and inline JavaScript.

  3. Deterministic scoring engine
    Findings set bit-flags → weighted penalties → a single final score with full traceability.

See TECHNICAL_OVERVIEW.md for detailed logic and implementation


Roadmap

Completed

  • Cookie bit-flag pyramid
  • CSP / header correlation
  • HTML <meta refresh> detection

In progress

  • Correlate CSP SHA to inline
  • Subresource-Integrity (SRI) hash checks
  • Consolidated CSP generator
  • Implement OpenSSL probe to retrieve certificate chain and reason for failed TLS handshake

License

GNU Affero GPL v3 – see LICENSE for details. Issues welcome.

Packages

No packages published