Suggestion: add WFGY 1.0 as a framework paper on AI infrastructure and self-healing LLM systems

[onestardao]

Hi, thanks for maintaining awesome-papers.
Your personal reading notes on machine learning systems, AI infrastructure, and related topics are very helpful.

I wanted to ask whether a framework-style paper on LLM self-healing and debugging could be added to this collection.

Project name: WFGY
Main paper DOI: 10.6084/m9.figshare.30338884
Main paper PDF: https://github.com/onestardao/WFGY/blob/main/I_am_not_lizardman/WFGY_All_Principles_Return_to_One_v1.0_PSBigBig_Public.pdf
GitHub repo: https://github.com/onestardao/WFGY

What WFGY 1.0 is

• A long-form framework paper that models large language models as self-healing systems, with a strong focus on system behavior.
• Describes modules and workflows for detecting, isolating, and repairing reasoning failures in production-style pipelines.
• Connects abstract formulas with practical prompt workflows that can be tried directly with existing LLMs.
• Published with a public DOI and supported by an open-source GitHub repository.

Why it might fit this project

• The repo already emphasizes machine learning systems and AI infrastructure, where reliability is a core concern.
• WFGY 1.0 focuses precisely on LLM reliability and self-healing at the system level, which matches that interest.
• It could be a useful addition for readers who want to think about debugging and robustness, not just model architectures.

Possible way to integrate

• Add WFGY 1.0 under a heading related to “AI systems / infrastructure / robustness”.
• Optionally include a short note that it is a framework paper with an accompanying open-source implementation.

If this does not fit how you want to curate awesome-papers, no problem at all.
Either way, thank you again for sharing these notes and references.

[NJX-njx]

I'd like to work on this. I'll add WFGY 1.0 to the paper list under a new LLM Reliability & Debugging section in paper-list/systems-for-ml/llm.md.

[mental2008]

Thanks for your suggestion! I will consider how to manage it in the current paper list.