📍 SuperQode v0.2.35: Poolside Laguna S 2.1 Local Support

Highlights

• Run Laguna S 2.1 using either DwarfStar or llama.cpp.
• Share one GGUF download between both inference engines.
• Resolve models from the standard Hugging Face cache without hardcoded user paths.
• Connect and launch Laguna directly from the SuperQode TUI.
• Use distinct DwarfStar variants:
  • Poolside Laguna S 2.1
  • Poolside Laguna S 2.1 Chat
  • Poolside Laguna S 2.1 Reasoner
• Open llama.cpp discovery in a dedicated model picker.
• Preserve Laguna reasoning and tool-calling behavior.
• Detect stale or incompatible DwarfStar builds.