note before publishing — these are my best guesses from folder

names and commit messages, not verified facts about your code. -->

OOPs-To-Agents

A staged progression from hand-rolled, OOP-based agents to full multi-agent frameworks.

---

What this is

A structured set of projects tracing one path through agent development: starting from plain object-oriented agent design, then layering in increasingly capable frameworks — CrewAI, LangChain, LangGraph — and finishing with a retrieval-augmented pipeline and an applied demo.

Each stage is kept as its own module so the jump between approaches — what a framework buys you over a hand-rolled implementation, and where that trade-off stops paying off — stays visible rather than getting flattened into one codebase.

Structure

Folder	Stage
Agents/	Framework-free agents built from base OOP principles — classes, inheritance, explicit state.
Crew/	Early CrewAI experiments.
CrewAi/	CrewAI-based multi-agent task orchestration.
Langchain/	LangChain agents and chains.
LangGraph/	Stateful, graph-based agent workflows in LangGraph.
RAG/	Retrieval-augmented generation pipeline experiments.
smart_energy_demo/	Applied capstone — an ACP/MCP-based smart-home energy agent.

Quickstart

git clone https://github.com/professorsab/OOPs-To-Agents.git
cd OOPs-To-Agents
pip install -r requirements.txt

Each folder is runnable independently — see the notes inside each one for specific entry points.

Status

Actively growing — each new folder represents a deliberate step up in framework capability rather than a replacement of the previous one.

License

MIT — see LICENSE.