agentscope-ai/agentscope-bricks ? reverse-engineered prompt

Build me a Python framework for making reusable AI agent components that I can plug into different agent apps.

I want to define small pieces like search, RAG, memory, and custom tools with clear input and output models, then reuse them with AgentScope, LangGraph, and AutoGen. It should feel production ready, with type safe schemas, automatic function style tool definitions, and easy adapters so the same component can work in different frameworks.

Please include a few simple examples and demos, plus enough tests and docs so I can understand how to make my own component and connect it to an agent. If any Alibaba Cloud or ModelStudio setup is needed, make that easy to configure with environment variables. Look up current docs online if you need to.