Оценка сложности портирования ML-модели

★ 6.5 · ai-tooling

analyze-model is a Claude Code skill that inspects an external machine learning model — typically hosted on Hugging Face — and estimates how difficult it would be to port into SD.Next using the port-model workflow. Given a model URL or Hugging Face repo id, the skill fetches repository artifacts such as model_index.json, config.json, and scheduler files, then classifies the model into one of five implementation buckets ranging from standard Diffusers to fully custom inference stacks. It produces a structured report covering classification with evidence, a difficulty rating on a Low/Medium/High