I kind of went down the rabbit hole on this one, it became a bit more than I had anticipated but I figured I’d share if anyone is interested.
I get frustrated that important details continually get dropped from my projects as I add other features or evolve other parts of a project when utilizing LLMs. So
I tried to create a workflow that adds specific agents to the task of referencing past knowledge and looping back in important context. I included everything I could find about the latest and most effective prompt engineering strategies.
This system is also self evolving and the agents are meant to improve over time as they learn from past experiences. I have it set to use API keys for the agents and utilize my postgres database.
(This is a placeholder, I’m on my phone and will upload file and GitHub for when I get home)