gemini-visionA
轻量 Gemini 多模态桥接，补齐 GLM / DeepSeek 等主模型的多模态能力缺口。分析视频/图片/音频 — 提炼参考风格、核验生成质量、对比版本差异、提取可复用框架。当用户需要"看视频""听音频""分析画面""检查生成质量""提炼参考风格""对比视频版本"时触发。Lightweight Gemini multimodal bridge for Claude Code — analyze video, images, and audio when the underlying model lacks multimodal input. Triggers on "watch/analyze this video", "check generation quality", "compare versions".
ai-agentsgobash
Skills Directory