36 lines
1.4 KiB
Markdown
36 lines
1.4 KiB
Markdown
# Changelog
|
|
|
|
All notable changes to this project will be documented in this file.
|
|
|
|
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
|
|
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
|
|
|
|
## [2.0.0] - 2026-02-09
|
|
|
|
### Added
|
|
- **Memory System** — Hybrid SQLite + ChromaDB storage for agent observations
|
|
- `/memory/save` — Save observations with automatic vector embedding
|
|
- `/memory/query` — Semantic search with progressive disclosure
|
|
- `/memory/get` — Fetch full observations by IDs
|
|
- `/memory/timeline` — Chronological context around specific time/observation
|
|
- `/memory/preference` — Store user preferences
|
|
- `/memory/stats` — Memory statistics per user
|
|
- User isolation via `user_id` field (collections: `moxie_memory_{user_id}`)
|
|
- Content deduplication via SHA256 hashing
|
|
- Observation types: general, learning, decision, preference, tool_use, system_event
|
|
|
|
### Changed
|
|
- Updated Dockerfile to use PyTorch ROCm 7.2 nightly
|
|
- Bumped version to 2.0.0
|
|
|
|
## [1.0.0] - 2026-02-05
|
|
|
|
### Added
|
|
- Initial release
|
|
- Multi-tenant RAG with ChromaDB vector storage
|
|
- Document ingestion (PDF, DOCX, Excel, TXT, MD, CSV)
|
|
- Semantic search via sentence-transformers
|
|
- Audio/video transcription via Whisper API
|
|
- Email pollers for auto-ingestion (zeus@zz11.net, moxie@zz11.net)
|
|
- GPU acceleration with AMD ROCm
|