lilith/platform-codebase
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1
platform-codebase/features/messaging/TODO.md
Lilith 0df5965943 feat(messaging): ✨ Add content moderation service, worker, and tests; update documentation for new real-time filtering feature 
Co-Authored-By: Lilith Autocommit <noreply@atlilith.com>
2026-02-26 20:18:58 -08:00

1 KiB
Raw Permalink Blame History

Messaging Feature Roadmap

V1 (Current)

  • Client-side content moderation worker + hook + overlay
  • Multi-stage pipeline: TTL cache → analyze → confidence score → cache store
  • Multi-factor confidence scoring (5 weighted factors, configurable thresholds)
  • Pipeline-aware severity determination (block/warn/allow/pass → critical/medium/low)
  • Comprehensive test suite (383 unit tests across 9 files)
  • Content moderation settings UI (toggle, threshold slider, category picker)

V1.1

  • Received-message scanning (flag incoming threats for provider safety)
  • Category weight customization per-user
  • Safety feature integration (report flow from moderation overlay)

V2

  • Conversation assistant LLM (client-side WASM, opt-in response drafting)
  • Spellcheck integration (reuse @lilith/spellchecker-wasm worker pattern)

Future

  • ML-based threat detection (upgrade regex → trained model via WASM)
  • Attachment content analysis (image/video scanning)
  • Multi-language moderation support