hemrajiscool/V2_CLI ? reverse-engineered prompt

I want you to build out this Python CLI project into a working version 2 for managing an AI helper with memory.

It should be able to take in text, store it locally, break it into chunks, create embeddings, and retrieve the most relevant past pieces when I ask a question. I also want a simple command line flow that feels easy to use, plus a small Telegram bot entry point if that fits with the current code.

Please make sure the app can use a local database and vector store for saved context, and wire up the main pieces so the ingest, retrieval, similarity search, and memory handling all work together cleanly. If anything needs current docs or best practice updates, look them up online and adapt the code.

Keep it practical, finish the rough edges, and make the whole thing feel like a usable v2 instead of a half done prototype.