Make PyFluent Architecture Agentic-Ready

[seanpearsonuk]

I. Agentic AI Platform

1. Make PyFluent Architecture Agentic-Ready

We’ll need solid scaffolding to make Fluent a safe and robust substrate for agentic workflows.
Key enablers include:

• Error transparency: Extend the exception hierarchy to support fine-grained error handling.
• Undo/redo: Explore lightweight transaction support via get_state() snapshots around set_state().
• Constraints: Expose enums and variable descriptors programmatically (ongoing). Investigate richer runtime constraints (valid ranges, dependencies).
• Observability: Enable tracing of natural-language → PyFluent calls → Fluent actions for explainability and debugging.
• Sandboxing: Investigate a future feature that would let users try out changes in a protected ‘test’ environment before anything becomes real. The system would check these trial changes for problems or inconsistencies, catching potential mistakes early and reducing the risk of breaking their actual setup.