-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Welcome to the Intent-Aware Security for Identity Verification documentation wiki.
This repository houses the source code and architectural documentation for a prototype "Privacy Wrapper" designed to secure identity verification systems (such as the Aadhaar authentication infrastructure). The system utilizes a dual-layer defense mechanism, combining Zero-Knowledge Proof (ZKP) cryptography with behavioral Machine Learning (ML) analysis, to detect anomalous behavior based on intent rather than relying solely on credential validity.
This wiki provides comprehensive technical documentation for all system components. Please use the sidebar or the links below to navigate through the topics:
-
System Architecture An overview of the middleware design, system boundaries, interaction flows, and the integration of frontend interfaces with the backend validation engine.
-
Cryptographic Security (ZKP) Detailed mathematical and implementation breakdown of the Zero-Knowledge Proof layer, including the Elliptic Curve Digital Signature Algorithm (ECDSA P-256) implementation, the challenge-response protocol, and replay-attack mitigation.
-
Machine Learning Engine Specifications regarding the behavioral analysis layer utilizing strictly unsupervised anomaly detection via the Isolation Forest algorithm. Contains details on feature selection, synthetic data distributions, and risk scoring logic.
-
API Reference Exhaustive documentation of the FastAPI endpoints, required JSON payloads, response structures, and HTTP status codes for system integration.
This project was developed as part of academic research under Group 4 - Introduction to Innovative Projects (PHY1901), focused on establishing robust, future-proof defenses against credential stuffing, automated scraping, and SS7/SIM swapping vulnerabilities inherent in legacy OTP-based systems.