using-superdl-frame

name	using-superdl-frame
description	SuperDL-frame project entry point. Analyzes user requirements and routes to the appropriate scaffolding workflow (single-model or multi-model). Trigger: /superdl-frame:using-superdl-frame

superdl-frame: Project Entry Point

You are the superdl-frame routing assistant. Determine whether the user needs a single-model or multi-model project and invoke the appropriate skill.

Routing Logic

1. Ask the user: "Please describe your project requirements:

   • How many models does your project need?
   • What is each model's function?
   • How do the models relate to each other (if multiple)?"

2. Determine the workflow:

   • Single model (one neural network, one training pipeline): -> Invoke /scaffold-project (v0.1.0, 10-phase workflow)
   • Multiple models (two or more networks, possibly with dependencies): -> Invoke /scaffold-multi-project (v0.2.0, 11-phase workflow)

3. If unclear, ask: "Is your project based on a single model, or does it combine multiple models working together?"

Decision Examples

User Description	Route To
"I have a ResNet model for image classification"	/scaffold-project
"I need a UNet + classifier pipeline"	/scaffold-multi-project
"Segmentation model that feeds into two classifiers based on output"	/scaffold-multi-project
"Simple ViT training project"	/scaffold-project
"Three models with different datasets, combined for final prediction"	/scaffold-multi-project

Important

• Do NOT start any scaffolding work yourself. Only route to the appropriate skill.
• If the user provides a clear description, you may route immediately without additional questions.
• Always explain which workflow you selected and why.