An LLM-powered knowledge extraction and evolution framework: transform unstructured documents into persistent, strongly-typed Knowledge Abstracts with one command — from simple collections and Pydantic models to knowledge graphs, hypergraphs, and even spatio-temporal graphs. Ships with 10+ extraction engines (GraphRAG, LightRAG, Hyper-RAG, KG-Gen, etc.), 80+ zero-code YAML templates, incremental evolution, and Obsidian export. It ranked 2nd on GitHub Trending worldwide.

The first hypergraph computation toolbox: a PyTorch-based deep learning library for graph and hypergraph neural networks, supporting both low-order and high-order message passing (vertex-to-vertex, vertex-to-hyperedge, hyperedge-to-vertex, vertex-set-to-vertex-set, etc.), with a built-in Auto-ML module for automatic hyperparameter tuning. It has been downloaded 50,000+ times and used by 200+ institutions across 24 countries.

A lightweight hypergraph database with built-in hypergraph visualization: native storage and querying of high-order relations, designed as the persistence layer for hypergraph-structured knowledge.

A self-consolidating semantic memory layer for AI agents: schema-first design with Pydantic, intelligent deduplication, and FAISS vector search — agent memory that organizes itself.

A flexible, AI-ready visualization engine for interactive graph/hypergraph exploration — bridging static graph visualization and dynamic AI-driven analysis with conversational interaction.

A pure-Python reader for KFB (KFBio) digital pathology slides with a fully OpenSlide-compatible API — drop-in replacement for OpenSlide on KFB whole-slide images, no native dependencies.

A pure-Python reader for MRXS (3DHISTECH MIRAX) whole-slide images with an OpenSlide-compatible API, enabling cross-platform access to MIRAX slides without vendor SDKs.

Whole-slide image alignment for H&E and multiplex IHC markers: a rigid + non-rigid registration pipeline that aligns multi-stain WSIs at gigapixel scale for spatially-resolved tumor microenvironment analysis.

Fast patch-based immunohistochemistry (IHC) inference on whole-slide images (SVS/KFB), powered by DeepLIIF — virtual IHC staining at WSI scale with batched patch inference.