[Feature] AgentVerse troubleshooting with the WFGY 16-problem RAG / agent map

[onestardao]

Hi, thanks for AgentVerse. It is a very flexible framework for multi agent experiments and simulations.

I work on WFGY, an open framework that defines a 16-problem failure map for RAG pipelines and agent systems:

• WFGY ProblemMap overview
  https://github.com/onestardao/WFGY/blob/main/ProblemMap/README.md

This map is referenced in:

• Harvard MIMS Lab ToolUniverse
• QCRI LLM Lab Multimodal RAG Survey
• University of Innsbruck Rankify project

Motivation

AgentVerse users often build large and complex scenarios. Common issues include:

• agents that appear rational locally but create unstable global behaviour
• very long dialogs where the original context is gradually lost
• confusion around shared memory or environment state.

These align with WFGY problems such as No.3 long chain drift, No.7 memory coherence, No.11 symbolic collapse and No.13 multi agent chaos.

Proposal

I would like to suggest a short troubleshooting guide in the AgentVerse docs that:

1. Uses the WFGY 16-problem map as a taxonomy for the most typical failures in AgentVerse experiments.
2. Provides a couple of small examples of how to label and analyse such failures.
3. Points readers to the full ProblemMap README for deeper reading.

If the maintainers are open to this idea, I can prepare a draft that follows the existing documentation style.