langchain-ai/langchain ? reverse-engineered prompt

Build me a Python framework that makes it easier to develop applications using large language models. The core idea is to provide a standard, modular way for developers to "chain" together different components. For example, it should be simple to connect to various LLM providers like OpenAI, Anthropic, or Google, and be able to swap them out a-la-carte without rewriting tons of code.

An important feature is connecting the models to private data sources for retrieval-augmented generation (RAG). I also want it to have powerful capabilities for building AI agents that can use tools to perform complex tasks. It should feel like a whole agent engineering platform. Let's make the architecture extensible, so the community can add new integrations easily. Maybe also include a separate, more advanced library for building controllable agent workflows that can handle loops and state.