Visual Assist is built with privacy as a core principle:

• On-Device Processing — All ML inference runs locally
• No Network Required — Works completely offline
• No Data Collection — Camera/audio data never leaves device
• No Analytics — Zero tracking or telemetry
• No Account Required — No user data stored

If you discover a security vulnerability, please report it responsibly:

1. Do not open a public issue
2. Email the author directly through the repository contact
3. Include detailed steps to reproduce the vulnerability
4. Allow reasonable time for a fix before public disclosure

• Description of the vulnerability
• Steps to reproduce
• Potential impact
• Suggested fix (if any)

Visual Assist follows these security practices:

• Minimum Permissions — Only requests necessary permissions
• No External Dependencies — Core functionality requires no third-party services
• App Transport Security — All network requests (if any) use HTTPS
• Keychain Storage — Sensitive data (if any) uses iOS Keychain

_{© 2026 Ayush. All rights reserved.}