Exposes a C++ codebase as a knowledge graph to AI agents for intelligent modernization tasks.
Sponsored
Legacy modernization of C++ codebases often fails due to lost context from "spaghetti code," AI agent hallucinations caused by unknown dependencies, and the inability of text-based parsers to understand C++ nuances. LegacyGraph solves these challenges by acting as a Model Context Protocol (MCP) server that transforms a C++ codebase into a queryable knowledge graph using accurate AST parsing. This allows AI agents to intelligently understand code structure, identify dependencies, detect circular dependencies, and prevent issues before refactoring, ensuring more reliable and accurate modernization.
Características Principales
010 GitHub stars
02Accurate C++ AST Parsing using tree-sitter
03Representation of C++ Codebase as a Knowledge Graph
04Detection of Circular Dependencies in the codebase
05Identification of Unused ("Orphan") Functions
06Universal Compatibility with various MCP clients (e.g., Claude Desktop, DeepSeek-Coder)
Casos de Uso
01Modernizing legacy C++ codebases with AI agents
02Identifying and resolving circular dependencies in complex C++ projects
03Discovering unused or "orphan" functions to clean up a codebase
