lattifai/awesome-embeddinggemma is a curated list repository that gathers local search applications, decision models, inference frameworks, tutorials, and datasets that use EmbeddingGemma2. The commit e8a6e7ccb684399c7bcbe2a49fd70c0e09fa9c60 of November 3, 2026 (UTC) added one new entry introducing semsearch to README.md and to the Simplified Chinese edition README.zh-CN.md respectively. The change is limited to those two files, with 2 insertions and 0 deletions.
The newly added target is spardanviro/semsearch. It is a local search tool that finds code semantically with EmbeddingGemma2 inside Claude Code; it uses syntax-aware chunking and an incremental SQLite index, and offers both a CLI and an MCP interface. Its runtime requirements are Python 3.12 and uv, plus a separate model download, and its inference targets CPU or NVIDIA CUDA. As the list description shows, the axis of this repository is not only search applications but also gathering decision models and datasets in one place. Like the other implementations in the same list, the new entry likewise requires downloading the model files separately before use.
That said, the content of the entry merely relays the descriptions given by each project, and is not a result of the list repository installing and verifying it directly. For semsearch, the author's test scope is Windows 11, while Linux and macOS have not yet been tested. The installation process touches user-level Claude Code settings and hooks, and if the retrieved code is passed to a cloud assistant, that snippet leaves the device. It appears to be a worthwhile option for people who want to add local search tied to Claude Code in a Windows environment, but support for other operating systems and the scope of the settings that installation changes can only be judged by checking the repository documentation directly. The source is the GitHub commit page: https://github.com/lattifai/awesome-EmbeddingGemma/commit/e8a6e7ccb684399c7bcbe2a49fd70c0e09fa9c60