Exposes rich semantic information from codebases to LLM code agents by integrating the Model Context Protocol (MCP) with the Language Server Protocol (LSP).
CodeSynapse is designed to provide a unified interface for querying semantic context from codebases, regardless of the underlying programming language. By integrating the Model Context Protocol (MCP) with the Language Server Protocol (LSP), CodeSynapse supports multiple languages through a configuration registry that maps file types to their respective language servers (e.g., Python files to Pyright, TypeScript files to tsserver), enabling LLM code agents to access rich semantic information in a consistent manner.
Key Features
01Maintains a configuration registry for language server mappings
02Supports multiple programming languages
030 GitHub stars
04Provides a unified interface for querying semantic context
05Integrates Model Context Protocol (MCP) and Language Server Protocol (LSP)
Use Cases
01Providing a consistent interface for code analysis across different languages
02Facilitating more intelligent code completion and suggestion features
03Enhancing LLM code agent capabilities with semantic code information
