Inspiration
Phishing emails increasingly look legitimate. We wanted to help users understand why a message is risky instead of giving them a black-box “safe” or “scam” label.
What it does
Inbox Guardian analyzes emails for suspicious senders, links, requests, and context. It provides a risk assessment, evidence ledger, recommended actions, and tools for detecting privacy leaks.
How we built it
We built a TypeScript web app combining deterministic safety checks with K2 Horizon. K2 reasons across sender identity, links, requests, urgency, and context to generate evidence-based findings.
Challenges we ran into
The main challenge was deciding what could be detected with rules versus what required contextual AI reasoning, while keeping every result understandable to the user.
Accomplishments that we're proud of
We built a working prototype that goes beyond a risk score by showing the evidence behind each assessment and providing actionable next steps.
What we learned
We learned that AI works best here as a reasoning layer alongside deterministic checks, especially when its conclusions are grounded in visible evidence.
What's next for Inbox Guardian
We plan to add deeper domain verification, attachment analysis, stronger privacy detection, and integration with real email workflows.
Built With
- api
- python