main-image-video
Skill71 starsChanged 3 months agoMITView on GitHub
静态主图转主图视频。一张商品图 → 3–5 秒可上架的主图短视频。当用户说「主图视频」「图转视频」「让图动起来」「加个视频」时使用。
In this repository25 other files it gives its agents.
- batch-imageSkill
- brand-kitSkill
- clothing-detailSkill
- clothing-extractionSkill
- clothing-grass-plantingSkill
- creative-sceneSkill
- cross-border-localizeSkill
- detect-taskSkill
- fabric-on-bodySkill
- fission-patternSkill
- flat-laySkill
- image-fusionSkill
- item-change-backgroundSkill
- item-detailSkill
- item-repairSkill
- item-selling-pointSkill
- listing-optimizerSkill
- material-enhancementSkill
- one-shotSkill
- platform-complianceSkill
- product-video-adSkill
- remove-watermarkSkill
- to-3dSkill
- ugc-testimonialSkill
- wear-everythingSkill
The file
Files
main-image-video — 主图视频
主图坑位旁边那个视频位,大多数店铺是空的。它不需要重新拍, 一张已有的主图就能生成。
一、能力边界
| 能做 | 说明 |
| 图生视频 | 一张商品图 → 3–5 秒运动镜头 |
| 运镜控制 | 推、拉、环绕、俯仰、微距扫过 |
| 材质动态 | 面料飘动、金属反光扫过、液体晃动 |
| 批量 | 配合 batch-image 整批出 |
| 不能做 | 说明 |
| 长视频 | 超过 8 秒请用 product-video-ad 分镜串联 |
| 口播 | 用 ugc-testimonial |
| 改商品 | 视频模型会放大原图的瑕疵,先用 item-repair 修好 |
二、先配后端
视频模型 ID 因后端而异,必须显式指定,脚本不会替你猜:
export DLAZY_VIDEO_MODEL=<你账号里可用的视频模型 ID>拼接与字幕需要 ffmpeg(本技能的单镜模式用不到,分镜才需要)。
完整说明见 references/video-backends.md。
三、工具调用
node scripts/video.mjs --mode clip --task main-image-video \
--image docs/flat-lay/example-output.jpg \
--prompt 'Slow push-in on the model. The cable-knit sweater fibers catch the light softly. Subtle natural body sway. Camera stays level, no cuts.' \
--seconds 3 \
--save out/sku001-main.mp4
# 先看会发什么,不计费
node scripts/video.mjs --mode clip --task main-image-video --image x.jpg \
--prompt '...' --dry-run参数约定
| 参数 | 取值 | 理由 |
--image | 一张主图 | 竖版 3:4 或方图,分辨率越高越好 |
--seconds | 3 ~ 5 | 主图位视频普遍偏短,超过 5 秒完播率掉得快 |
--brand | brand.yaml | 让运镜与色调跟全店一致 |
四、Prompt 模板
结构:运镜 + 主体动态 + 环境动态 + 约束
<Camera move>. <What the product does>. <What the environment does>.
Keep the product identical to the reference image — same shape, color, texture, logo placement.
No cuts, no text overlay, no watermark. Photorealistic.运镜词表
| 目标 | 英文 |
| 缓推 | slow push-in |
| 缓拉 | slow pull-back reveal |
| 环绕 | smooth 30-degree orbit around the product |
| 俯冲 | gentle top-down tilt |
| 微距扫过 | macro pan across the surface texture |
材质动态词表
| 品类 | 英文 |
| 针织 / 毛呢 | fibers catch the light, fabric breathes subtly |
| 真丝 / 雪纺 | fabric ripples in a gentle draft |
| 金属 / 珠宝 | specular highlight sweeps across the metal |
| 玻璃 / 液体 | liquid settles, light refracts through the glass |
| 皮革 | soft sheen shifts across the grain |
五、执行流程
- 先检查静态图。视频模型会放大原图的一切瑕疵。原图有崩就先修,别指望视频救。
- 确认后端。
DLAZY_VIDEO_MODEL没设就先问用户用哪个模型。 - 单张试跑。视频比图贵得多,先跑一条 3 秒的看运镜对不对。
- 确认后再批量。
- 过一遍平台规格。视频也有规格要求(时长、比例、体积), 目前 platform-compliance 只校验静态图,视频请人工对照平台文档。
六、常见问题
| 现象 | 原因 | 怎么办 |
| 商品在动的过程中变形 | 运动幅度给太大 | 换成 slow / subtle 系的词,缩短时长 |
| 画面糊 | 原图分辨率不够 | 先用 material-enhancement 提清晰度 |
| 模特脸在动的时候崩 | 人脸是视频模型的重灾区 | 改用不露脸的景别,或改成商品特写 |
| 时长对不上 | 多数模型的时长是档位不是连续值 | 按模型支持的档位取,别硬凑 |
| 报错说没指定模型 | DLAZY_VIDEO_MODEL 没设 | 见「先配后端」 |
Tips
- 3 秒够用。主图位视频不是广告片,讲清楚一个卖点就行。
- 别加文字。主图视频的文字规则和主图一样严,字幕留给 product-video-ad。
- 首帧就是封面。用商品最好看的角度做输入图,因为它大概率是静止时显示的那一帧。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

