DeusData/codebase-memory-mcp

High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

What it solves

AI coding agents often struggle to understand large codebases because they rely on file-by-file exploration or simple grep searches, which consume massive amounts of tokens and require many tool calls. codebase-memory-mcp provides a high-performance, local knowledge graph of a codebase, allowing agents to perform structural queries and impact analysis with significantly fewer tokens and higher accuracy.

How it works

The engine uses tree-sitter AST analysis across 162 languages and a "Hybrid LSP" for semantic type resolution in popular languages (like Python, TypeScript, and Rust) to build a persistent knowledge graph of functions, classes, and call chains. It is implemented as a native executable (written in C) that uses a RAM-first pipeline with LZ4 compression and in-memory SQLite for extreme indexing speed. It integrates with AI agents via the Model Context Protocol (MCP), providing 17 specialized tools for search and analysis.

Who it’s for

Developers using AI coding agents (such as Claude Code or Codex) who need the agent to have a deep, structural understanding of large-scale repositories without the overhead of slow indexing or high token costs.

Highlights

  • Extreme Performance: Indexes the Linux kernel (28M LOC) in 3 minutes and answers structural queries in under 1ms.
  • Token Efficiency: Reduces token usage by up to 120x compared to file-by-file exploration.
  • LSP-Enhanced: Supports semantic type resolution for 12 major languages to produce high-fidelity call graphs.
  • Built-in Visualization: Includes a 3D interactive graph UI served directly from the binary at localhost:9749.
  • Local & Private: All processing happens 100% locally with no requirement for API keys, Docker, or external hosted services.
  • Infrastructure-Aware: Indexes Dockerfiles and Kubernetes manifests as graph nodes with cross-references.
  • Broad Language Support: Supports 162 languages via vendored tree-sitter grammars.

Related

  • Project
    johnhuang316/code-index-mcp

    An MCP server that provides intelligent code indexing and AST-based analysis to help AI assistants navigate and understand complex codebases.

  • Project
    aovestdipaperino/tokensave

    tokensave is a Rust‑based local server that builds a semantic knowledge graph of a codebase, letting AI coding agents query symbols, relationships, and snippets in a single call. By replacing costly grep/glob scans, it cuts token usage, supports multi‑branch indexing, offers crash‑resilient syncing, and integrates with dozens of LLM‑powered coding tools.

  • Project
    forloopcodes/contextplus

    An MCP server that turns large codebases into searchable, hierarchical feature graphs using RAG, AST parsing, and spectral clustering for high-accuracy AI coding context.

  • Project
    jgravelle/jcodemunch-mcp

    An MCP server that uses tree-sitter AST parsing to let AI agents retrieve precise code symbols instead of entire files, reducing token costs by up to 99%.

  • Project
    colbymchenry/codegraph

    CodeGraph builds a local, up‑to‑date knowledge graph of a codebase (30+ languages) and plugs into AI coding assistants (Claude, Cursor, Copilot, etc.). By providing agents with a single, precise payload of symbols and call‑paths, it cuts tool calls, token usage and latency—benchmarks show up to 44 % cost reduction and 88 % fewer tool calls.