A Chrome extension that analyzes every domain you visit and shows a real-time trust score — powered by AI on Cloudflare's global edge network.
Built by Keerti Panchakshari Charantimath · LinkedIn
Trust scores surface right where you need them — no extra steps required.
Open the browser extension on your current site to check its safety.
Right-click any link to check its safety before navigating to the page.
Every domain runs the same three-stage pipeline, cached at the edge for speed.
Domain captured from the active tab and sent to POST /analyze.
① Heuristics — brand, TLD, entropy → ② Workers AI (Llama 3.1 8B) → ③ Community votes, weighted.
Cached? Return instantly. New? Run the pipeline and store the result for 6 hours.
● 75–100 Trusted · ● 42–57 Caution · ● 0–21 Dangerous
The average phishing site is live for 21 hours before any blocklist catches it.
Safe Browsing only catches reported domains — hours after they're live.
Registering g00gle.xyz takes seconds. Most users won't notice.
A page that loads clean HTML and renders a phishing form via JavaScript is invisible to proxies.
No tool combines AI + community signals at the browser layer, in real time, at the moment you click.
Especially useful for high-value targets and security teams.
Gets phishing emails mimicking internal tools. Needs protection that requires no behavior change.
Wants a browser-layer signal to act on — warn or isolate — within their existing policy engine.
Wants a fast check on links from Slack, GitHub, or email before navigating.
Each feature closes a specific gap.
Color-coded score updates automatically on every navigation. Green = trusted, red = dangerous.
Amber banner injected at the top of the page for caution-level domains. Dismissible, auto-disappears after 10s.
Click the shield icon for the complete breakdown — trust gauge, per-signal scores, plain-English AI explanation.
Vote Safe / Suspicious / Unsafe from the popup. Weighted votes from real users feed back into scoring.
Flag wrong verdicts directly from the popup. Creates a feedback loop that improves accuracy over time.
Community votes correct AI mistakes — especially on brand-new domains with no threat-feed history.
AI + heuristics produce initial verdict
Safe, Suspicious, or Unsafe
Active reviewers carry up to 2×
60%+ majority escalates or de-escalates
Architecturally ready — requiring engineering effort to ship.
