Features Deep Dive

Code-y is packed with features to provide a complete code understanding platform. Here’s a closer look at what it can do.

🧠 Advanced Code Analysis Engine

At its core, Code-y uses deep AST (Abstract Syntax Tree) parsing for both TypeScript and JavaScript.

Entity Recognition: Automatically identifies React Components, functions, classes, types, and custom hooks.
Relationship Mapping: Tracks imports, component usage, method calls, and inheritance to build a complete picture of your codebase.
Advanced Detection: Automatically flags circular dependencies and potential prop drilling.
Intelligent Duplicate Detection: Finds exact, near, and structural duplicates to help you refactor and reduce redundancy.

Leverage the power of AI, all running on your local machine for maximum privacy.

Flexible AI Backends: Full integration with Ollama for 100% local processing. Optional support for the OpenAI API if you provide a key.
Vector Similarity System: Uses a local FAISS vector database and the Nomic Embed Text model to enable powerful semantic code search.
AI-Generated Documentation: Automatically generate context-aware descriptions for components, functions, and props.

The generated web UI is a powerful tool for exploring your code.

Multi-View Dashboard: Dedicated explorers for components, functions, and classes.
Advanced Visualization: An interactive, force-directed relationship graph lets you visually navigate your code's architecture.
Semantic Search: Go beyond keyword search to find code based on its meaning and purpose.
Code Quality Analysis: View detailed reports on code duplication and other architectural insights.

Chat with an AI assistant that has full context of your codebase.

Codebase-Aware Conversations: Ask questions in natural language like "How does the authentication flow work?".
Refactoring Suggestions: Get AI-powered recommendations for improving your code.
Session Memory: The chat maintains context, allowing for follow-up questions and deeper exploration.