Skip to content

Releases: tanishqmudaliar/SilverGuard

v1.0.0 — First Stable Release: Licensed & Fully Documented

Choose a tag to compare

@tanishqmudaliar tanishqmudaliar released this 16 Jul 13:34

SilverGuard's first stable, public-ready release. The project is now MIT licensed and fully documented for open-source use and portfolio presentation.

Added

  • MIT License
  • Complete README rewrite covering architecture, AI/ML pipeline, threat classification, database schema, privacy & security notes, build configuration, and troubleshooting

Notes

This marks SilverGuard as ready for public use, contribution, and showcase.

Full Diff: v0.2.1...v1.0.0

v0.2.1 — Monitoring & UI Responsiveness Patch

Choose a tag to compare

@tanishqmudaliar tanishqmudaliar released this 16 Jul 13:34

A refinement release focused on SMS monitoring reliability and UI polish.

Changed

  • Improved real-time SMS monitoring with better data handling
  • Added dedicated scroll controllers for smoother, more responsive UI
  • Minor Android build configuration adjustments

Full Diff: v0.2.0...v0.2.1

v0.2.0 — Guardian Protection: Contact Alerts & Notifications

Choose a tag to compare

@tanishqmudaliar tanishqmudaliar released this 16 Jul 13:33

This release introduces the guardian system — allowing SilverGuard to notify trusted contacts when a scam or suspicious SMS is detected.

Added

  • Guardian contact management (add, edit, remove trusted contacts)
  • Notification service with actionable Dismiss and Report buttons
  • SMS sender service to alert guardians directly via text
  • New Settings page for managing guardians and preferences

Changed

  • Extended the local database schema to support guardian records
  • Updated SMS and contacts services to integrate with the guardian alert workflow

Full Diff: v0.1.0...v0.2.0

v0.1.0 — Initial Release: On-Device Scam Detection Core

Choose a tag to compare

@tanishqmudaliar tanishqmudaliar released this 16 Jul 13:33

Initial release of SilverGuard — an AI-powered Android SMS scam detection app built for protecting users, especially senior citizens, from fraudulent text messages using on-device machine learning.

Added

  • Real-time SMS monitoring using Android's SMS_RECEIVED broadcast via another_telephony
  • On-device scam classification powered by a fine-tuned MobileBERT model running through ONNX Runtime
  • SQLite database layer for persisting scanned and classified messages
  • Background scam processor for asynchronous, non-blocking message analysis
  • Contacts and permission handling services
  • Documentation for external ONNX model hosting on Hugging Face (MODEL_HOSTING.md)

Full Diff: https://github.com/tanishqmudaliar/SilverGuard/commits/v0.1.0