SECURITY
PROJECT NOTES / 2025–2026PhishGuard.
A phishing-awareness prototype with model-assisted explanations, practice scenarios, and a learning dashboard.
Build the habit
of looking closer.
Pause.
Question.
Recognise.
The input is text, rather than a fetched page or file contents.
An explanation to help someone think through the warning signs.
Learning history helps shape future exercises.
A text-based advisory and practice workflow. The visual represents learning, not a verified detection result.
Original diagram · based on public source01 / THE CONTRIBUTION
My part in the project.
Repository owner, with public application and API code for phishing-awareness exercises and explanations.
02 / INSIDE THE BUILD
The choices underneath.
Connect threat explanations with practice scenarios so the interface supports learning as well as inspection.
Use server-side model calls for scan explanations, chat, and scenario generation.
Store learning context through Supabase to support personalised practice.
03 / STATUS & LIMITS
Where it stands.
- Scan explanations are based on URL text or filenames; the model does not fetch the URL or inspect file contents.
- Confidence labels are generated by the model and have not been calibrated as detection probabilities.
04 / KEEP EXPLORING
Read it at the source.
Explore the code and follow the latest state of the project in its repository.
View PhishGuard on GitHubAbout these project notes
Based on the public scan, scenario-generation, dashboard, and Supabase integration source.