This handles image analysis through MiniMax's vision MCP tool. It auto-triggers when you reference image files or ask to analyze, describe, or extract text from screenshots, mockups, charts, or photos. You get five analysis modes: general description, OCR for text extraction, UI review with design critique, chart data extraction, and object detection. Requires a MiniMax Token Plan subscription and MCP setup. The skill includes environment-specific config examples for OpenCode, Claude Code, and Cursor, plus mode-specific prompts that shape how the vision model interprets your image. Supports common formats up to 20MB and handles both URLs and local paths.
npx skills add https://github.com/minimax-ai/skills --skill vision-analysis