Role naming incompatibility when switching between Claude and Gemini agents

[ololoDevs]

I've encountered an issue with conversation history when switching LLM providers between agents in Neuron AI.
The problem:
When an agent uses Claude, the conversation history stores AI responses with the role "assistant". However, when I switch that same agent to use Gemini, the responses are stored with the role "model".
The error:
When I switch back to Claude (or use Claude in a subsequent agent), it throws an error because Claude doesn't recognize the "model" role in the conversation history. Claude API only accepts "user" and "assistant" roles according to Anthropic's documentation.
Example flow:

Agent A uses Claude → history contains "assistant" messages ✅
Switch Agent A to Gemini → new messages stored as "model" ✅
Switch Agent A back to Claude → Error: Claude doesn't understand "model" role ❌

My questions:

Is this a known issue in Neuron AI framework?
Does Neuron AI automatically normalize/translate role names when switching between different LLM providers?
If not, what's the recommended approach to handle this? Should I:

Manually transform the conversation history?
Clear history when switching models?
Use separate conversation contexts per provider?

Are there plans to implement automatic role name translation/normalization across different LLM providers in future versions?

This makes it difficult to dynamically switch between models within the same conversation flow, which would be a valuable feature for cost optimization and capability selection.
Thanks for your help!

[ilvalerione]

Fixed in 2.8.5