Dsh Vision Skill

Key features

• (c) View (runtime skill)Models load as much as possible as a graphic instruction.
• Qwen Official Dynamic Resolution Pre-Processing is headed, Smart resize is responsible for the pixel budget and patch grid adsorption, which is then handed over to any OpenAI compatible multimodular model.
• It has packaged the warehouse General skills/Vision-skill (from Qwen ' s official dynamic resolution method) into a DSH raw plugin that allows pure text models that do not support pictures to see maps, OCRs, positioning targets.
• A progressive tool is exposed.
• Auto-activate the following tools after loading (progressive exposure)