Hey everyone, I am new to the community and am working on an agent factory project. Essentially, I am designing a autonomous agentic feedback loop that takes in a standardised survey with informastion about the requirements the agent is expected to fulfill for any given business or application, and the agentic loop will generate a full conversational flow agent with tools, a KB file, as well as simulation tests, and a ton of local verification to ensure the agentic loop doesn't output any invalid structures. I am having trouble finding a standard or best-practices docuement to add to my project for the agent to reference. Does anyone know if such a document exists? I have it creating agents already though the api, but I am having trouble documenting the best practices for prompt writing inside of subagent nodes, when to use an extract variable node, when to use a function node, etc. Thanks!