Prakni by Kanthira Labs

Phishing attacks have evolved. Your defenses should too.

Prakni is a privacy-first security extension that acts as your personal cybersecurity analyst inside Gmail. It scans your inbox for sophisticated scams, explains exactly why an email is dangerous, and stops you from clicking malicious links — all without a single byte of your email ever leaving your computer.

🛡 WHAT PRAKNI DOES

● Detects Brand Impersonation: Catches scammers pretending to be your bank, PayPal, or government services.
● Uncovers Hidden Links: Spots typosquats, invisible redirects, and fake domains (e.g., paypa1.com instead of paypal.com).
● Spots Social Engineering: Flags emails that combine manufactured urgency with financial requests.
● Active Link Shield: If you accidentally click a malicious link, Prakni intercepts the click and throws up a warning screen before you reach the fake website.
● Explains the Threat: We don't just block emails. We tell you exactly WHY an email was flagged so you can learn what to look for.

🔒 100% PRIVACY BY DESIGN

Security shouldn't cost you your privacy.

● Local Analysis Only: Your email content, subject lines, and sender names are analyzed entirely inside your browser.
● Zero Cloud Processing: We never transmit, read, or store your emails on our servers.
● No Tracking: No analytics, no ads, and no data selling. Ever.

Privacy Policy: https://kanthiralabs.github.io/prakni-privacy/

⚡ SMART & SILENT

● Tuned for Real Threats: Specifically built to catch hyper-local scams like fake GST notices, KYC panics, and tax refund fraud that global filters miss.
● No False Alarms: Prakni stays completely silent when your mail is safe. You only see banners when there is a real threat.
● Always Evolving: Threat rules update seamlessly in the background so you are always protected from the latest scams.

⚙ FULLY CUSTOMIZABLE

● Take Control: Use the built-in dashboard to manage your Trusted Domains and Safe Contacts.
● Adjust Sensitivity: Choose between Conservative, Balanced, or Aggressive scanning modes.
● Mark Safe: False alarm? One click whitelists the sender and instantly removes the warning.

Built by KanthiraLabs.
All analysis is local. Always.