Find your landing page message
Connect a product’s value to a specific audience and write a clear first-screen message.
WeRead-inspired HyperFrames video template for vertical annual reading reports, personal reading dashboards, book-note recaps, and shareable year-in-review stories. Use when users want a 9:16 HTML-to-MP4 reading report with warm paper texture, editorial Chinese typography, book-page metaphors, da...
---
name: weread-year-in-review-video-template
description: |
WeRead-inspired HyperFrames video template for vertical annual reading reports,
personal reading dashboards, book-note recaps, and shareable year-in-review
stories. Use when users want a 9:16 HTML-to-MP4 reading report with warm paper
texture, editorial Chinese typography, book-page metaphors, data highlights,
and deterministic motion.
triggers:
- "WeRead year in review"
- "WeRead annual report"
- "reading year in review video"
- "annual reading report template"
- "微信读书年度报告"
- "读书年度总结视频"
- "阅读年报 HyperFrames"
od:
mode: template
surface: video
type: hyperframes
platform: mobile
preview:
type: html
entry: example.html
reload: debounce-100
design_system:
requires: false
outputs:
primary: index.html
secondary:
- template.html
- example.html
example_prompt: "Create a WeRead-style 9:16 HyperFrames annual reading report video with 12 scenes, warm paper texture, book-page transitions, reading stats, notes, keywords, and a final reading persona card."
capabilities_required:
- file_write
---
# WeRead Year in Review Video Template
Create a vertical HyperFrames composition for annual reading reports: WeRead,
Goodreads, Readwise, Notion reading logs, book clubs, or personal learning
recaps. The template turns reading time, active days, bookshelf assets, notes,
keywords, and a reading persona into a shareable 9:16 video.
## Resource Map
```text
weread-year-in-review-video-template/
├── SKILL.md
├── assets/
│ ├── template.html
│ └── default-showcase.mp4
├── references/
│ └── checklist.md
└── example.html
```
## Workflow
1. Copy `assets/template.html` to `index.html`.
2. Replace the default report data in the `REPORT` object:
- owner/title
- reading hours and active days
- bookshelf and completion stats
- note composition
- interest keywords
- reading persona and share line
3. Preserve the 12-scene timeline unless the user asks for a shorter cut.
4. Keep the WeRead-inspired visual language:
- warm paper background
- ink-blue typography
- restrained WeRead green accents
- book pages, bookmarks, highlights, note cards, and shelf metaphors
5. Motion should feel like flipping through a reading journal. Avoid techy
slide transitions, bouncy UI effects, and dashboard-loading motion.
6. Keep the composition deterministic:
- direct `data-start`, `data-duration`, and `data-track-index` attributes
- no unseeded randomness
- no infinite loops or `repeat: -1`
- no dependency on scroll, hover, localStorage, or runtime class discovery
7. Validate against `references/checklist.md` before emitting.
## Output Contract
Emit one short orientation sentence, then a single HTML artifact:
```xml
<artifact identifier="weread-year-in-review-video-template" type="text/html" title="WeRead Year in Review Video Template">
<!doctype html>
<html>...</html>
</artifact>
```
---
name: weread-year-in-review-video-template
description: |
WeRead-inspired HyperFrames video template for vertical annual reading reports,
personal reading dashboards, book-note recaps, and shareable year-in-review
stories. Use when users want a 9:16 HTML-to-MP4 reading report with warm paper
texture, editorial Chinese typography, book-page metaphors, data highlights,
and deterministic motion.
triggers:
- "WeRead year in review"
- "WeRead annual report"
- "reading year in review video"
- "annual reading report template"
- "微信读书年度报告"
- "读书年度总结视频"
- "阅读年报 HyperFrames"
od:
mode: template
surface: video
type: hyperframes
platform: mobile
preview:
type: html
entry: example.html
reload: debounce-100
design_system:
requires: false
outputs:
primary: index.html
secondary:
- template.html
- example.html
example_prompt: "Create a WeRead-style 9:16 HyperFrames annual reading report video with 12 scenes, warm paper texture, book-page transitions, reading stats, notes, keywords, and a final reading persona card."
capabilities_required:
- file_write
---
# WeRead Year in Review Video Template
Create a vertical HyperFrames composition for annual reading reports: WeRead,
Goodreads, Readwise, Notion reading logs, book clubs, or personal learning
recaps. The template turns reading time, active days, bookshelf assets, notes,
keywords, and a reading persona into a shareable 9:16 video.
## Resource Map
```text
weread-year-in-review-video-template/
├── SKILL.md
├── assets/
│ ├── template.html
│ └── default-showcase.mp4
├── references/
│ └── checklist.md
└── example.html
```
## Workflow
1. Copy `assets/template.html` to `index.html`.
2. Replace the default report data in the `REPORT` object:
- owner/title
- reading hours and active days
- bookshelf and completion stats
- note composition
- interest keywords
- reading persona and share line
3. Preserve the 12-scene timeline unless the user asks for a shorter cut.
4. Keep the WeRead-inspired visual language:
- warm paper background
- ink-blue typography
- restrained WeRead green accents
- book pages, bookmarks, highlights, note cards, and shelf metaphors
5. Motion should feel like flipping through a reading journal. Avoid techy
slide transitions, bouncy UI effects, and dashboard-loading motion.
6. Keep the composition deterministic:
- direct `data-start`, `data-duration`, and `data-track-index` attributes
- no unseeded randomness
- no infinite loops or `repeat: -1`
- no dependency on scroll, hover, localStorage, or runtime class discovery
7. Validate against `references/checklist.md` before emitting.
## Output Contract
Emit one short orientation sentence, then a single HTML artifact:
```xml
<artifact identifier="weread-year-in-review-video-template" type="text/html" title="WeRead Year in Review Video Template">
<!doctype html>
<html>...</html>
</artifact>
```
Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.