跳到正文
VibeFather
免费注册

用这个模板做的 2 个案例

同一份提示词,不同的 [VARIABLES] 变量。点一个在上方预览。

关于这个模板

弹幕视频平台是一个受 Bilibili 启发的媒体社区,围绕观众反应成为影片一部分的那一刻构建。它包含可搜索的首页信息流、六张原创视频缩略图,以及一个观看视图,其中代码绘制的 Canvas 场景在定时评论下方播放。密度图揭示对话聚集的位置,而播放、评论筛选、收藏、关注和分享都无需服务器即可响应。

视觉系统采用深墨色表面、暖琥珀色点缀、紧凑的元数据和清晰的 Space Grotesk 与 Inter 字体。在首页,一部大型精选影片和一个实时对话面板立即确立产品定位。观看页保持播放器主导,相关影片近在咫尺,上传者卡片、评论编辑器和本地活动状态增添上下文。评论进入真实轨道,而不是作为装饰性覆盖层出现。在较小屏幕上,相同的控件会重排为适合触摸的行,减少动态效果设置会暂停场景,直到观看者选择播放。

此模板适合创作者、社区建设者以及探索交互媒体或社交播放的开发者。更改品牌、标语、内容焦点、受众和精选标题,以将目录适配到另一个编辑主题。保留包含的图片并重写周围的标题,以符合每个画面展示的内容,或用同样连贯的一组替换本地资源。第二个案例展示了一个使用相同布局和交互的旅行与美食出版物。

一切作为静态 Vite 和 React 构建运行。用户评论、收藏影片、关注和观看状态保留在浏览器中,使模板适用于原型、作品集项目和客户端产品演示,而无需暴露凭证或要求外部视频服务。

模板提示词

在线演示就是用这份提示词原样做出来的,没有额外补充。其他案例只是改了 [VARIABLES] 变量的值。 提示词保持英文原版,国内常用的 AI 编程工具都能直接理解;下面附中文参考译文。

# Danmaku Video Platform

Build this website exactly as described below. Replace the [VARIABLES] with your own values; keep everything else.

## Brief
Build [BRAND], an original English-language danmaku video platform for [AUDIENCE], focused on [CONTENT_FOCUS]. Its editorial line is [TAGLINE]. Include a searchable home feed and a complete watch experience with a code-drawn Canvas film, synchronized lane comments, a density timeline, comment filters, local comment injection, an uploader card, and related films. The default featured film is [FEATURED_VIDEO].

- Audience: The default audience is [AUDIENCE]: people who enjoy carefully made visual films and want to react at a precise moment. The template also serves creators and developers who need a convincing client-side social video product rather than a static media-gallery mockup.
- Goal: Let a visitor discover a film, start playback immediately, understand where conversation peaks occur, filter the overlay, seek through the timeline, and add a comment that appears in the stream and density map. Make saving, following, sharing, search, sorting, and related-video navigation work locally.
- Tone: Warm, observant, and culturally curious rather than loud or juvenile. Use compact, specific editorial copy and believable viewer reactions. The interface should feel energetic because the conversation moves, while the underlying layout stays calm and readable.

## Stack
- Framework: Vite 5 + React 18 + TypeScript
- Styling: Tailwind CSS 3.4
- Libraries: lucide-react (ISC), @fontsource/space-grotesk (OFL), @fontsource/inter (OFL), @fontsource/jetbrains-mono (OFL)
- Build: `npm run build` → static files in `dist/`. No backend, no server code, no external APIs.
- Setup:
  - Set Vite base: './' and build.sourcemap: false. Implement the two views with hash routes '#/' and '#/watch' using React state and window.location.hash; do not add a server or router dependency.
  - Keep all variable values, video records, categories, timed comments, uploader data, and interface copy in src/content.ts. Rewrite default-specific copy when variables change, derive handles and internal identifiers from [BRAND], and keep the six image subjects aligned with their rewritten titles.
  - Import only the listed Space Grotesk, Inter, and JetBrains Mono weights from their Fontsource packages so every font file is bundled locally.
  - Use a devicePixelRatio-aware Canvas 2D renderer capped at 2x, requestAnimationFrame for playback, transform-only danmaku movement, and full cleanup of timers and listeners on route or video changes.
  - Use Lucide icons for controls and navigation. Give icon-only buttons accessible names, visible keyboard focus, and minimum 44px touch targets on mobile.
  - Persist saved video IDs, followed channel IDs, locally added comments, and comment preferences in localStorage. Keep playback time in memory and provide a safe fallback when storage or clipboard access is unavailable.

## Variables
- [BRAND] — Platform brand (default: "Rillcast"). Invented mascot-free video-community name.
- [TAGLINE] — Editorial tagline (default: "Watch together. Think out loud."). One short line used in the featured area and footer.
- [CONTENT_FOCUS] — Catalogue focus (default: "visual culture, design, and urban stories"). The editorial theme used to rewrite film descriptions and collections.
- [AUDIENCE] — Audience (default: "curious makers and visual learners"). The people the catalogue and footer address.
- [FEATURED_VIDEO] — Featured film title (default: "A City Before It Wakes"). Title of the first film; keep it compatible with the dawn-city image.
- [CREDIT] — keep a small "Template by VibeFather" link to https://vibefather.ai/t/danmaku-video-platform in the footer (default: yes; delete this line to remove it)

The copy under Pages & Sections is written for the default values. If your values differ, rewrite every headline, paragraph, label and image for them — same length, tone and structure. Layout, colors, type and motion stay as described.
Keep all copy, image paths and links in one content file (for example `src/content.ts`), keyed by section, so the content can be swapped later without touching components.

Pro 模板

解锁完整提示词和源码

  • 完整提示词:设计、版块、动效、素材
  • 每个在线演示的源码
  • 商用授权

在线演示和 DESIGN.md 仍然免费。

$9/月起 · 年付和终身正在上线特惠

DESIGN.md

单独的设计规范:配色、字体、间距和动效。可以把这套视觉用到你已有的项目上。

---
name: "Danmaku Video Platform"
colors:
  canvas: "#0B0D12"
  surface: "#141821"
  raised: "#1B2230"
  line: "#2B3444"
  text: "#F5F2E9"
  muted: "#9AA5B6"
  amber: "#F2C14E"
  coral: "#F27D62"
typography:
  display_film_titles_and_wordmark: { family: "Space Grotesk", weights: [500, 600, 700] }
  interface_descriptions_and_comments: { family: "Inter", weights: [400, 500, 600] }
  timecodes_counts_and_density_labels: { family: "JetBrains Mono", weights: [400, 500] }
spacing: "Use a 4px base unit. Cap the main shell at 1500px with 40px desktop gutters, 24px tablet gutters, and 16px mobile gutters. Header height is 72px desktop and 64px mobile. Use 24px card padding, 32px section gaps, and 48px between the featured area and feed; keep dense metadata at 8px line gaps."
radius: "Use 18px for media cards and the player, 14px for panels, 10px for buttons and inputs, and full pills only for compact filters and status badges. Media corners stay crisp; use 1px Line borders instead of heavy shadows."
---

# Danmaku Video Platform — Design System

Apply this design system to an existing project. It describes the look and feel only; for the full site, use prompt.md.

## Colors
- **Canvas** `#0B0D12` — Page background and player letterbox
- **Surface** `#141821` — Cards, navigation, and conversation panel
- **Raised** `#1B2230` — Inputs, hovered cards, and control menus
- **Line** `#2B3444` — One-pixel borders and inactive timeline bars
- **Text** `#F5F2E9` — Primary text and high-contrast danmaku
- **Muted** `#9AA5B6` — Metadata, timestamps, and secondary labels
- **Amber** `#F2C14E` — Primary actions, active states, playhead, and density peaks
- **Coral** `#F27D62` — Creator comments and small conversational highlights

## Typography
- **Display, film titles, and wordmark** — Space Grotesk (OFL), weights 500/600/700. Self-host the font files.
- **Interface, descriptions, and comments** — Inter (OFL), weights 400/500/600. Self-host the font files.
- **Timecodes, counts, and density labels** — JetBrains Mono (OFL), weights 400/500. Self-host the font files.

## Spacing & shape
- Spacing: Use a 4px base unit. Cap the main shell at 1500px with 40px desktop gutters, 24px tablet gutters, and 16px mobile gutters. Header height is 72px desktop and 64px mobile. Use 24px card padding, 32px section gaps, and 48px between the featured area and feed; keep dense metadata at 8px line gaps.
- Radius: Use 18px for media cards and the player, 14px for panels, 10px for buttons and inputs, and full pills only for compact filters and status badges. Media corners stay crisp; use 1px Line borders instead of heavy shadows.

## Motion
- Render each film from normalized playback time so seeking produces the same frame immediately; cap scene animation at 30fps while keeping controls responsive.
- Move danmaku with horizontal transforms in collision-aware lanes. Pause freezes every comment, seeking re-evaluates active comments, and changing speed adjusts comment travel consistently.
- Use 160ms view fades, 180ms card-image scale from 1 to 1.025, 140ms button state changes, and a 220ms drawer slide. Never animate large blur filters.
- With prefers-reduced-motion, load the player paused on a composed frame, disable automatic comment motion and preview loops, remove reveal transforms, and allow explicit manual playback.

## Imagery
Use the six listed generated editorial photographs only as local 16:9 thumbnails and featured-media artwork. Their consistent petrol-blue shadows, amber light, and fine grain are the photographic signature. The watch player must instead render original animated scenes in Canvas 2D; never place a static thumbnail behind fake playback. Use abstract code-drawn channel monograms rather than portraits, mascots, anime, or recognizable people.

## Tone
Warm, observant, and culturally curious rather than loud or juvenile. Use compact, specific editorial copy and believable viewer reactions. The interface should feel energetic because the conversation moves, while the underlying layout stays calm and readable.

---
— Prompt by VibeFather · https://vibefather.ai/t/danmaku-video-platform · CC BY-NC 4.0 (non-commercial; commercial use needs a VibeFather Pro license) · Request removal: https://vibefather.ai/takedown

改成你自己的

  • 一个电影社团:将 BRAND 设置为你集体的名称,将 CONTENT_FOCUS 指向艺术影院,并围绕你的放映活动重写六个影片标题和上传者卡片。
  • 一个设计工作室作品集:将 AUDIENCE 用于潜在客户,将 FEATURED_VIDEO 设置为一个旗舰项目,并将相关影片替换为案例研究短片。
  • 一个音乐厂牌:将 TAGLINE 改为你的艺人阵容标语,将密度图评论替换为曲目反应,并将上传者卡片链接到你的艺人页面。
  • 一个课堂媒体频道:将 AUDIENCE 设置为学生,保留评论筛选用于讨论管理,并围绕讲座影片重写首页信息流。

弹幕视频平台:常见问题

弹幕视频平台模板中的定时评论如何与视频保持同步?

弹幕视频平台模板播放的是代码绘制的 Canvas 场景而非视频文件,每条评论都带有时间戳。播放期间,场景通过 requestAnimationFrame 推进,时间戳落入当前窗口的评论会在真实轨道上移动,而不是作为装饰性覆盖层停留。拖动时间轴会重新定位流,密度图会更新以显示反应聚集的位置。

怎么用 AI 做一个弹幕视频平台?

复制本页的完整提示词(Pro),或者运行 npx vibefather add danmaku-video-platform,把 [BRAND]、[TAGLINE]、[AUDIENCE]、[CONTENT_FOCUS] 换成你自己的内容,再粘贴到 Claude Code、Codex、Cursor 等 AI 编程工具里。在线演示就是用这份提示词原样做出来的。