[GH-ISSUE #472] Suggestion: add WFGY as a long-horizon LLM debugging app #449

New Issue

GiteaMirror · 2026-04-15T06:53:44-05:00

GiteaMirror commented

2026-04-15 06:53:44 -05:00

Originally created by @onestardao on GitHub (Feb 7, 2026).
Original GitHub issue: https://github.com/Shubhamsaboo/awesome-llm-apps/issues/472

Hi Shubham,

Thanks for maintaining this list, it has become a default place I check when I look for serious LLM apps and agent projects. It is very helpful to see everything curated in one place.

I maintain an open source project called WFGY, which might fit here as a slightly unusual kind of LLM app.

Project: WFGY
Repo: https://github.com/onestardao/WFGY

Very short description:

WFGY is an open source framework for debugging and stress testing large language models under long horizon, high tension text scenarios.

WFGY 1.0 is a PDF that introduces the conceptual framework and treats LLMs as self-repairing systems.
WFGY 2.0 focuses on real world failure modes such as RAG pipelines, vector stores, and deployment ordering.
WFGY 3.0 Singularity Demo is a single TXT pack that behaves like a debugging app: you upload the TXT to a model, follow a menu, and it walks the model through a sequence of high-tension questions so you can see where its reasoning starts to drift or collapse.

From the point of view of this repo, WFGY is essentially an “LLM app for testing other LLM apps.” It is not another agent framework, but rather a long-horizon workload you can point any agent or chat system at to audit its behavior over time.

If you feel this is in scope, a possible place could be under the more advanced LLM apps, or in a small subsection for evaluation and debugging tools. I am happy to send a small pull request in whatever format you prefer. If you think it does not really fit the focus of this list, no worries at all, thanks for considering and for keeping this resource up to date.

Originally created by @onestardao on GitHub (Feb 7, 2026). Original GitHub issue: https://github.com/Shubhamsaboo/awesome-llm-apps/issues/472 Hi Shubham, Thanks for maintaining this list, it has become a default place I check when I look for serious LLM apps and agent projects. It is very helpful to see everything curated in one place. I maintain an open source project called WFGY, which might fit here as a slightly unusual kind of LLM app. Project: WFGY Repo: https://github.com/onestardao/WFGY Very short description: WFGY is an open source framework for debugging and stress testing large language models under long horizon, high tension text scenarios. WFGY 1.0 is a PDF that introduces the conceptual framework and treats LLMs as self-repairing systems. WFGY 2.0 focuses on real world failure modes such as RAG pipelines, vector stores, and deployment ordering. WFGY 3.0 Singularity Demo is a single TXT pack that behaves like a debugging app: you upload the TXT to a model, follow a menu, and it walks the model through a sequence of high-tension questions so you can see where its reasoning starts to drift or collapse. From the point of view of this repo, WFGY is essentially an “LLM app for testing other LLM apps.” It is not another agent framework, but rather a long-horizon workload you can point any agent or chat system at to audit its behavior over time. If you feel this is in scope, a possible place could be under the more advanced LLM apps, or in a small subsection for evaluation and debugging tools. I am happy to send a small pull request in whatever format you prefer. If you think it does not really fit the focus of this list, no worries at all, thanks for considering and for keeping this resource up to date.

GiteaMirror closed this issue

2026-04-15 06:53:44 -05:00

GiteaMirror commented

2026-04-15 06:53:45 -05:00

@awesomekoder commented on GitHub (Feb 8, 2026):

Hi @onestardao, thanks for the suggestion! However, this repo focuses on self-contained tutorials that teach developers how to build AI/LLM applications — it's not a directory of external tools. If you'd like to contribute, consider creating a tutorial that teaches the concepts behind WFGY with code included in this repo. Thanks for understanding!

@awesomekoder commented on GitHub (Feb 8, 2026): Hi @onestardao, thanks for the suggestion! However, this repo focuses on self-contained tutorials that teach developers how to build AI/LLM applications — it's not a directory of external tools. If you'd like to contribute, consider creating a tutorial that teaches the concepts behind WFGY with code included in this repo. Thanks for understanding!

GiteaMirror commented

2026-04-15 06:53:45 -05:00

@onestardao commented on GitHub (Feb 9, 2026):

Yes I already sent the PR , and do the tutorials, the 16 problem map is useful that my stars almost all comes from that, thanks ^^

@onestardao commented on GitHub (Feb 9, 2026): Yes I already sent the PR , and do the tutorials, the 16 problem map is useful that my stars almost all comes from that, thanks ^^