Developed int8 support for HGraph within the antgroup/vsag repository, enabling memory-efficient indexing for small numeric datasets. The work involved implementing algorithmic changes in C++ to handle int8 data types, updating build scripts, and providing a clear usage example that demonstrates tangible performance improvements in vector search scenarios. By introducing a dedicated demo file and example-driven documentation, the developer facilitated easier adoption of efficient data structures for users. The technical approach focused on algorithm implementation, index building, and data structure optimization, with all changes thoroughly integrated and documented. No major bugs were reported, reflecting a focused and well-executed feature delivery.