‹ 首页

muapi-product-ad-cinematic

@samuraigpt · 收录于 1 周前 · 上游提交 3 周前

Cinematic 5–10s product ad from a product photo + brand brief.

适合你,如果你需要快速生成高质量的产品广告视频。

/ 下载安装
muapi-product-ad-cinematic.skill双击,或拖进 Claude 桌面版 / Cowork,即完成安装↓ .skill↓ .zip
用别的 agent?下载 .zip 解压,把文件夹放进它的技能目录
Claude Code~/.claude/skills/(项目级 .claude/skills/)
Codex CLI~/.codex/skills/
Cursor自动读取上面两处目录
其他工具见其文档的「skills」目录;两个下载是同一份文件,只是名字不同
/ 通过 npx 安装 校验哈希
npx oh-my-skill add samuraigpt/generative-media-skills/muapi-product-ad-cinematic
/ 通过 bash 安装
curl -fsSL https://oh-my-skill.com/install.sh | bash -s -- samuraigpt/generative-media-skills/muapi-product-ad-cinematic
/ 已经装过?验证本机副本,不用重装
npx oh-my-skill verify samuraigpt/generative-media-skills/muapi-product-ad-cinematic
安装目标可用 --agent / --scope 或 --to 明确指定;省略时只会在唯一已存在的 agent 目录上自动选择,零命中或多命中会停止并提示。content_hash 缺失或不一致均拒装。
3754GitHub stars
~713上下文体积 · 单文件
镜像托管

怎么用

商店整理自技能原文 · 版本 a1c4c98 · 表述以原文为准
它做什么

装上后,Claude 会根据你提供的产品照片和品牌风格描述,先生成 4 张不同风格的静态画面供你挑选,再把你选中的画面放大、制作成 5–10 秒的电影感视频,并配上背景音乐。

什么时候触发

当你提到“产品广告”、“商业广告”、“电影感广告”或“产品视频”等关键词,或者直接要求为产品生成一段电影感视频时触发。

装好后可以这样说
Claude 会生成 4 张画面供你挑选,然后制作视频。
Claude 会生成 4 张画面供你挑选,然后制作视频。
技能原文 SKILL.md作者撰写 · MIT · a1c4c98

Cinematic Product Ad

Cinematic 5–10s product ad from a product photo + brand brief.

Inputs

| Name | Type | Required | Default | Description | |:---|:---|:---|:---|:---| | product_image | image_url | yes | — | URL of the product photo (must already be uploaded). | | brand_brief | text | yes | — | Mood / style direction (e.g. "luxury minimal", "playful"). | | duration_sec | int | no | 6 | Final video length in seconds (5–10). |

Steps

This skill has TWO phases separated by a user pick. Submit them as two separate the plan calls — never bundle downstream steps into the first plan.

Phase A — variant exploration (cheap)

Submit ONE the plan containing only:

  1. Hero frame variants — 4 separate muapi image generate nodes (model=nano-banana-2, aspect_ratio=16:9 by default).
  2. Each prompt restyles the product against the brand brief mood. Vary lighting, palette, framing, and lens between variants. Keep product geometry intact.
  3. Reference the user's product_image if the model supports image conditioning; otherwise describe the product in detail.

After the plan executes, end your turn with a brief message listing the 4 asset_ids and asking the user which one to take forward (e.g. "Pick a hero (asset_1, asset_2, asset_3, or asset_4)?"). Wait.

Phase B — commit on the picked hero (expensive)

Once the user replies with their pick, submit a SECOND the plan:

  1. Upscale the picked frame — enhance_image (operation=upscale).
  2. Animate the upscaled frame — muapi video from-image (model=kling-v3.0-standard-image-to-video, duration={{duration_sec}}, prompt="slow cinematic push-in, soft volumetric light, subtle product micro-rotation"). Reference the upscale's URL with $nX.url.
  3. Background musicmuapi audio create (kind=music) — runs in parallel with the upscale/animate. Style derived from brand_brief (luxury → "ambient cinematic, warm strings, slow tempo, instrumental"). Duration ≈ video length.
  4. Return the upscaled hero image and the final video.
Notes
  • If the brief mentions "luxury", bias the palette to gold/black; for "playful", bias to bright/saturated.
  • If video gen fails after failover, fall back to a still-frame slideshow (just return the upscaled hero + music).
  • Don't auto-confirm step 4 — its cost (~80 cr) deserves a user nod.
Trigger Keywords

product ad, commercial, cinematic ad, product video


Notes for the Executing Agent
  • This recipe is LLM-orchestrated: read each phase, gather any missing inputs from the user, then call muapi CLI commands. Use muapi auth configure first if MUAPI_API_KEY is unset.
  • For model IDs without a CLI alias yet, fall back to the raw endpoint via curl -X POST https://api.muapi.ai/api/v1/<endpoint> -H "x-api-key: $MUAPI_API_KEY" -H 'content-type: application/json' -d '{...}' and poll with muapi predict wait <request_id>.
  • Substitute {{input_name}} placeholders with the user's actual inputs before issuing each call.
按 MIT 许可原样转载,未经改动 · 在 GitHub 查看 →

评论

登录即可评论;带「已验证安装」的,是发布者名下有本店的安装或持有记录。