Tokennuke is an MCP (Model Context Protocol) server focused on intelligent code indexing, exposing 15 tools for AI assistants and agents. It enables instant, efficient code symbol access, hybrid keyword+vector search, function/call graph mapping, and remote repository indexing, supporting 10 major languages via tree-sitter. Designed for code intelligence use cases—AI coding agents, code review, semantic code search, refactoring, and onboarding—it integrates smoothly with MCP/AI agent clients like Claude, Cursor, and Windsurf. Systems/clients can interact with Tokennuke via stdio or HTTP; it provides O(1) symbol retrieval, incremental indexing, and efficient token usage for code navigation and annotation.
Visit Tokennuke's official website for product details and getting started.