After seeing sophisticated phishing attempts slip through Gmail's filters, a friend and I built a browser extension that uses local AI to analyse emails for threats.
It runs entirely in your browser using web-llm, with any threat being clearly labelled with an explanation of why it's suspicious.
Still early days, but we'd love feedback from the HN community. Even if you're not comfortable installing it yet, we have screenshots showing how it works.
https://chromewebstore.google.com/detail/eguard/dnpofinmilhb...