v0.0.9
✨ New Features
- Knowledge Evolution Framework: Introduced a signal-driven engine that automatically organizes, merges, and refines the runtime knowledge graph after each search.
- Meta-Cluster Routing: Optimized similarity retrieval using a community-based directory layer, significantly reducing per-query computation.
- Graph Visualization: Added a new D3.js force-directed knowledge cluster graph page to the Web UI, complete with sidebar links and i18n support.
🚀 Enhancements
- Storage Upgrades: Added joint cluster+embedding reads, richer merge semantics, and reverse-edge cleanup on deletion.
- Non-Blocking Execution: Knowledge evolution now runs asynchronously (fire-and-forget) with reentrancy locking, LLM concurrency control, and a kill switch to guarantee zero impact on search latency.
🐛 Bug Fixes
- Fixed Python syntax errors (missing commas) in README code examples.
- Resolved D3 graph runtime
TypeErrors by adding null guards and safe assertions. - Fixed Next.js hydration mismatch when rendering localized timestamps.
📖 Documentation
- Updated
README.mdandREADME_zh.mdto detail the new runtime knowledge evolution engine.
What's Changed
- [Feature] Introduce Knowledge Evolver Module by @fanqiNO1 in #170
- [Feature] Introduce Knowledge Graph Visualization by @fanqiNO1 in #171
New Contributors
Full Changelog: v0.0.8...v0.0.9
Contributors
fanqiNO1