Hi, and thank you for creating Learn Prompting. It is one of the few repos that really treats prompt engineering as a learning path, not only a bag of tricks.
I maintain an open source project called WFGY, and inside it there is a standalone page named “WFGY 16 Problem Map”.
Link: https://github.com/onestardao/WFGY/tree/main/ProblemMap/README.md
What students get from it:
- A compact list of 16 failure modes that frequently appear when they start building real RAG or LLM powered apps.
- Each item has:
- typical “symptoms” (what they observe in the chat or tool),
- a minimal repro pattern,
- and structural fixes (for example change chunking contract, adjust vectorstore settings, add eval, etc).
- The language is practical and provider agnostic, so students can apply it with any stack.
Possible way to integrate:
- A short link in the section where you talk about RAG, evaluation, or troubleshooting.
- It can be positioned as “a checklist to use when things keep breaking even after prompt tuning”.
If it does not fit your style or syllabus, no worries at all. Either way, thank you for making Learn Prompting available to everyone.
Hi, and thank you for creating Learn Prompting. It is one of the few repos that really treats prompt engineering as a learning path, not only a bag of tricks.
I maintain an open source project called WFGY, and inside it there is a standalone page named “WFGY 16 Problem Map”.
Link: https://github.com/onestardao/WFGY/tree/main/ProblemMap/README.md
What students get from it:
Possible way to integrate:
If it does not fit your style or syllabus, no worries at all. Either way, thank you for making Learn Prompting available to everyone.