Skip to content

Suggestion: add WFGY 3.0 · open source LLM debugging toolkit (16 Problem Map) #3

Description

@onestardao

Hi, thanks for putting this awesome AI toolkit together.

I would like to suggest adding WFGY as a developer-facing toolkit for debugging and stress-testing LLM applications:

Very short description:

  • An open source framework that uses a 16 Problem Map of failure modes to debug RAG and LLM apps.
  • Includes concrete checklists for vector stores, retrieval logic, prompt wiring, and long-horizon reasoning.
  • Designed so that a developer can paste their failing prompt / code context and quickly see which problem type they are hitting, plus suggested fixes.

Why it might fit this list:

  • It is not a general “AI toy”, it is specifically built for LLM reliability engineering.
  • Pairs well with other tools in this repo that focus on training and deployment, by covering the “why is my LLM app broken in production” side.
  • Fully open source, already used in the wild as a free “clinic” for RAG and LLM bugs.

If you feel it fits, I am happy to send a PR and place it in a section like “LLM tools for developers” or “Debugging and evaluation”.

Thanks for considering it.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions