Tuning AI Agents with Repo Rules and Living Docs
Treat custom rules and living docs as a RAG simulation: retrieve the right context fast so an agent solutions against your decisions instead of grepping at random.
Ends with a final quiz. Score 70% or better and you can download a completion certificate in your own name — shareable on LinkedIn, verifiable by its certificate code.
Get access
This course is free. Contribute whatever you'd like to support future courses — or leave it at 0 and enroll now.
A practical course on tuning coding agents with repo rules and living documents. You will design a small trusted corpus (rules, a what-doc, a how-doc), simulate retrieval without a vector store, write a ranker you can test, and see how stale docs produce confident wrong edits. Includes an in-page compiler lab, token-budget and assembly visualizers, and a stale-document trap.