MCP.so
登录
C

Codegraph Rust MCP Server

@Jakedismo

基本信息

分类

其他

传输方式

stdio

作者

Jakedismo

提交者

Saijani Super

关于 Codegraph Rust MCP Server

🎯 Overview

Codegraph Rust MCP Server 配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "codegraph": {
      "command": "codegraph",
      "args": [
        "start",
        "stdio"
      ],
      "env": {
        "CODEGRAPH_CONFIG": "~/.codegraph/config.toml"
      }
    }
  }
}

Codegraph Rust MCP Server 工具

工具列表暂未提供

当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Codegraph Rust?

Codegraph Rust is a CLI tool that combines MCP (Model Context Protocol) server management with advanced code analysis. It indexes projects, manages embeddings, and runs MCP servers with multiple transport options, supporting languages like Rust, Python, JavaScript, TypeScript, Go, Java, and C++.

How to use Codegraph Rust?

Install and run via CLI. Use commands for project indexing, MCP server management (STDIO, HTTP, or both simultaneously), code search (semantic, exact, fuzzy, regex, AST), and architecture analysis. Specific configuration options include embedding models, performance tuning, and background daemon mode with PID management.

Key features of Codegraph Rust

  • Multi-language code parsing with Tree-sitter
  • Dual transport support: STDIO and HTTP streaming
  • Vector search with FAISS-powered embeddings
  • Graph-based architecture with RocksDB storage
  • High performance: 170K lines in 0.49 seconds
  • Incremental indexing with file watching

Use cases of Codegraph Rust

  • Index large codebases and perform semantic search across multiple languages
  • Run MCP servers in STDIO, HTTP, or dual mode for flexible agent integration
  • Analyze component relationships, dependencies, and code patterns
  • Build a deep code knowledge synthesis system with an agent

常见问题

1

What languages does Codegraph Rust support?

It supports Rust, Python, JavaScript, TypeScript, Go, Java, and C++.

2

What transport options are available for MCP servers?

STDIO, HTTP streaming with SSE, or both simultaneously. Background daemon mode with PID management is also supported.

3

How fast is code indexing?

Parsing 170K lines of Rust code takes about 0.49 seconds; generating 21,024 embeddings takes about 3 minutes 24 seconds on an M3 Pro with Qdrant/all-MiniLM-L6-v2-onnx (CPU, no Metal acceleration).

4

Does Codegraph Rust support incremental indexing?

Yes, it supports incremental indexing with file watching and smart caching.

5

What kind of code search is available?

Semantic search using embeddings, exact match, fuzzy search, regex, and AST-based queries with configurable similarity thresholds.

评论

其他 分类下的更多 MCP 服务器