Skill Claude CodeCodex
Use this skill whenever the user shares an image, asks about an image ("这是什么"、"他是谁"、"图片里有什么"、"看图"), or mentions an image file path. The current model has NO native vision — images arrive as "Unsupported Image" — so you MUST use the llm-vision MCP's describeimage tool to actually see it. For identifying a person or…
0 27d ago A 0 tokens