code-memory
MCP server with local vector search for your codebase. Smart indexing, semantic search, Git history — all offline.
code-memory is a deterministic, high-precision code intelligence layer exposed as a Model Context Protocol (MCP) server, designed to provide accurate and token-efficient code context for LLMs. It runs entirely locally, indexing codebases for semantic search, architecture understanding, and Git history analysis, addressing the cost and inaccuracy of traditional context retrieval methods.
Features
Compatibility
Quick start
Use cases
Alternatives
Related searches
Comments
- PParker NguyenMay 25, 2026
Finds code by what it does, not just what it's named — that's the differentiator.
- JJordan ChenMar 29, 2026
Works well for large codebases where grep falls short.
- FFinley JohnsonMar 24, 2026
Semantic search over your codebase that actually finds relevant code.
- RRowan NguyenMar 14, 2026
Indexing is fast enough to stay current with active development.
- CCasey JohnsonMar 12, 2026
Local vector search means no data leaves your machine.