aws-samples/amazon-bedrock-workshop ? reverse-engineered prompt

I want to build a hands on workshop environment for learning Amazon Bedrock. It should be organized into separate modules covering the main Bedrock capabilities: calling foundation models directly through inference APIs, building knowledge bases with RAG so models can answer questions from my own documents, customizing models with fine tuning, and creating AI agents that can use tools and take actions.

Make it all run through interactive Jupyter notebooks so people can actually execute code and see results as they learn. Each module should have clear examples they can run cell by cell. Set it up with proper Python dependency management, maybe using uv since that's modern and fast.

The workshop needs to handle prerequisites like AWS credentials, IAM permissions for Bedrock and related services like OpenSearch for the RAG stuff and DynamoDB for agents. Include setup instructions assuming people might be running this locally or in a provided AWS environment. Make sure there are warnings about costs since running foundation models isn't free.

Keep the notebooks educational with explanations mixed in with the code. This is for developers and solution builders who want to actually understand how to use Bedrock, not just copy paste. Structure the whole thing so someone can clone the repo, install dependencies, and start working through progressively more advanced AI capabilities.