DOCTRINE · ACTIVEv2026.04 · Security Architecture
Design Philosophy & Safety

TRUSTED
INQUIRY

“How are learning records protected?” “How is data kept safe?” We balance the freedom of inquiry-based learning with high security standards.

LEARNING CONTINUITY
Personalized inquiry

Saving the inquiry process (history)

Conversations with AI and terrain you focused on are saved as an inquiry trail, so you can resume seamlessly from where you left off.

AI tuning (System Instruction)

Gemini instructions adapt to learning level and interests, so feedback stays deeper and more relevant.

User Experience

"From one-off experiences to continuous learning."

SECURITY & CREDITS

Transport protection (TLS/SSL)

End-to-end encryption over HTTPS helps keep learning data from being intercepted in transit.

XSS defenses

Escaping and sanitization at render time help block injection of malicious scripts.

API key protection

Keys are managed via environment variables and are not exposed on the client side.

CSP configuration

Script execution is restricted at the browser level to trusted sources.

Data Source / Map data attribution

This site uses Geospatial Information Authority of Japan (GSI) map tiles and map data from MIERUNE (OpenStreetMap contributors).

Backend outlook

We plan secure authentication with Firebase Auth and Google OAuth without storing passwords ourselves. Row-level security (RLS) on the database isolates and protects each person’s inquiry data.

Firebase AuthPostgres RLSTLS 1.3CSP Ready