AI Native Landscape

Codebase Memory MCP

Tracked

High-performance code intelligence MCP server that indexes codebases into a persistent knowledge graph. Supports 158 languages with sub-millisecond queries and 99% fewer tokens, shipped as a single static binary with zero dependencies.

Author DeusData Open Sourced 2026-02-24 Last Commit Unknown

Overview

Codebase Memory MCP is a high-performance code intelligence server built on the Model Context Protocol. It parses a codebase into a persistent knowledge graph using tree-sitter and exposes sub-millisecond structural queries, giving coding agents precise, token-efficient access to symbols, call graphs, and definitions across 158 languages. Delivered as a single static binary with no runtime dependencies, it drops straight into Claude Code, Codex, Cursor, Windsurf, Aider, Gemini CLI, and other MCP-compatible clients.

Key Features

  • Persistent code knowledge graph indexed via tree-sitter, covering 158 programming languages with AST-level symbol and relationship extraction.
  • Sub-millisecond query latency with Cypher and graph-traversal interfaces for symbol lookups, call graphs, and dependency paths.
  • Token-efficient retrieval that returns only relevant structural slices, cutting context costs by roughly 99% versus loading whole files.
  • Single static binary, zero runtime dependencies — deployable locally or in CI with no install chain.
  • Broad agent compatibility, integrating with Claude Code, Codex, Cursor, Windsurf, Aider, Gemini CLI, Kilocode, and OpenCode.
  • Built-in graph visualization for exploring code relationships and architecture interactively.

Use Cases

  • Giving coding agents accurate, repo-wide structural context without flooding the prompt with raw source.
  • Large-codebase navigation, impact analysis, and "who calls what" exploration during refactors or reviews.
  • Onboarding and architecture understanding through visual graph exploration of symbols and dependencies.
  • CI and automation pipelines that need fast, deterministic code queries without a heavy indexing stack.

Technical Details

  • Model Context Protocol (MCP) server exposing code-intelligence tools to any compatible agent client.
  • Storage backed by SQLite with a graph schema, enabling durable indexing and offline-friendly local operation.
  • tree-sitter drives multi-language AST parsing; results are queryable via Cypher and graph APIs.
  • Average repository indexes in milliseconds, with sub-millisecond query response for symbol and relationship lookups.