Ztrimus/text-to-visuals

AI Summary: Integrate AWS Bedrock foundational models into a text-to-visuals application using LangChain. This involves creating a modular script that isolates model logic, provides a simple interface for other backend components, and adheres to the existing architecture (documented in docs/architecture.md and server/model_bedrock.py). The goal is to enable the application to use Bedrock models for intent understanding and intermediate representation (IR) generation to create visual outputs from textual input.