-
Notifications
You must be signed in to change notification settings - Fork 407
Description
Hi, thanks a lot for creating and maintaining this course – it’s one of the clearest introductions to prompt engineering I’ve seen.
I’d like to suggest one external resource as an optional advanced reference for students who later start building RAG or tool-using systems and hit reliability issues.
Resource
- Name: WFGY 16 Problem Map
- Link: https://github.com/onestardao/WFGY/blob/main/ProblemMap/README.md
What it is
A concise map of 16 common failure modes in real-world LLM systems, especially when you go beyond “pure prompting” and start using:
- RAG over your own documents
- vector databases and custom chunking
- longer multi-step chains and agents
Each problem entry includes:
- a short description in plain language
- what it looks like from the user’s point of view
- typical root causes (retrieval, chunking, indexing, prompts, etc.)
- suggested fixes and debugging steps
The idea is to give people a checklist they can walk through when “the prompt looks fine, but the system still fails.”
Why it might fit this repo
The course already teaches people how to design good prompts.
When they apply those prompts inside RAG / LangChain / agent setups, they will inevitably hit system-level bugs.
Linking to the Problem Map (perhaps in a “Further reading / RAG & reliability” section) could give motivated learners a structured way to:
- see typical failure patterns before they hit them
- debug their first real projects more systematically
Possible way to add
For example, near the end of the README under a small “Further reading” section:
- WFGY 16 Problem Map – 16 common failure modes in RAG / LLM systems with debugging tips and mitigation patterns.
If this doesn’t fit the scope you have in mind for the repo, no problem at all – thank you again for the excellent material.