⚡

CodeMind

Semantic code intelligence
Online

Upload. Index. Ask anything.

Drop your repository ZIP, explore with AI-powered retrieval

About this project

CodeMind (AI Debug) is a retrieval-augmented generation (RAG) system for codebases. Upload a ZIP of your source code, and it indexes functions and files into a semantic search store. Ask natural-language questions — the system retrieves the most relevant code chunks and uses Google Gemini to explain how your project works.

Built with FastAPI, sentence-transformers embeddings, intent-aware retrieval, and a session-scoped in-memory vector index. Your Gemini API key stays in your browser and is only sent per query — never stored on the server.

View source on GitHub anikaitar70/aidebug Built by Anikaitar github.com/anikaitar70
1 Upload codebase
Indexing repository 0%
Processed 0 / 0 files
Chunks created 0
2 Ask your codebase
How your key is handled
  • Saved in your browser only (sessionStorage)
  • Sent to our server only when you run a query so we can call Gemini on your behalf
  • Never stored in our database, session index, or server logs

Project overview

◎

Upload a repository to begin