LiveCanvas skill
A skill that turns a topic into a short animated card with facts from found sources and Live Photo assets for the iPhone
- 539 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The agent searches the web and runs code: it installs npm dependencies and starts a render
- Python and FFmpeg scripts write files into the task directory
- Data from found pages ends up in the finished card and needs checking against sources
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/livecanvasDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent searches the web and runs code: it installs npm dependencies and starts a render
- Python and FFmpeg scripts write files into the task directory
- Data from found pages ends up in the finished card and needs checking against sources
This is third-party code. Review the repository files before installing.
What it does
LiveCanvas turns a single sentence into an information card: a cover and a short animation that explains the topic. The skill first searches the web for facts and saves the source data and references, and only then writes headlines and captions. If no suitable time series exists, it builds the card from facts, a process or a scene and does not invent data to fit a chart. The agent then drafts a short plan: the main line of the set, the takeaway of each card, a storyboard and the asset plan. The frame holds objects from the topic (a product image, a coffee cup, a train, a medal, a cityscape), emoji, a timeline and a chart when needed. The animation is built in Remotion and by default is 1080×1440, 30 frames per second, about three seconds and silent. All of these values can change. The output is a cover, an MP4, a list of data sources and an editable Remotion project. If the machine has Python 3.10 or newer, Pillow and FFmpeg, a .pvt package with a photo and video pair for Live Photo is added. The skill supports its old name LiveChart: a $livechart call is redirected to LiveCanvas if the livechart folder sits next to it.
Who it is for. For content managers, marketers and editors who make short information cards and want the agent to verify the numbers and build the animation itself.
Good fit when
- You need an animated card with the dynamics of a metric, such as a year of temperatures in two cities
- You want to explain a news item or product release as a short scene with numbers and links to sources
- You need a comparison of two products or metrics as a cover and a three-second clip
- You need Live Photo assets to put the animation on an iPhone screen
- You want an editable Remotion project to tweak the animation by hand
Not a fit when
- You need clips longer than a few seconds or with sound: the tool is built for silent cards
- The agent has no internet and no code execution: the skill searches the web and renders through commands
- Node.js, Python and FFmpeg are not installed: the animation cannot be built without them
- You need ready footage for editing on a long timeline: this makes short self-contained cards
Example request
Use LiveCanvas to make a card about temperature changes in Moscow and Sochi over the last full year: a cover, a three-second animation and a list of sourcesLimitations
The animation needs Node.js with npm, Python 3, FFmpeg and suitable fonts, and project dependencies are installed with npm ci. Live Photo on Linux, Windows and macOS needs Python 3.10 or newer, Pillow and FFmpeg, and macOS also has a Swift path with PhotoKit verification. The tool currently produces silent H.264 video and a JPEG photo.
The .pvt package holds the photo, video and metadata, but the Files app on iPhone may not preview it. The project description shows animations as GIF files, which are not a native Live Photo. The skill does not import the result into the library itself.
The template is based on Remotion 4.0.530. Remotion ships under its own license, which differs from MIT, so read its terms on the Remotion site before commercial use. The sample data in the template is synthetic and good only for testing, so verify facts against sources before publishing. The skill searches the web for data, so the quality of a card depends on the sources found.
How to disable. Delete the livecanvas folder, and livechart if present, from the agent's skills directory, for example `~/.codex/skills/`. Work results live in the task directory and are deleted by hand.
FAQ
How do I install LiveCanvas in Claude Code or Codex?
Run `npx skills add pengchujin/livecanvas --skill livecanvas -g` and pick your assistant. For Codex only, add the `-a codex` flag. Manually, put the whole `livecanvas/` folder into `~/.codex/skills/`; SKILL.md alone is not enough.
What is a .pvt file and how do I put the animation on an iPhone?
It is a package with a photo and video pair and metadata for Live Photo. The skill creates it when the machine has the needed Python, Pillow and FFmpeg. A preview of the package in the Files app on iPhone may not appear.
Where does LiveCanvas get the numbers for a card?
The skill searches the web and saves the source data and references before writing headlines. For time series there is a `validate_data.py` script that checks dates, source references and coverage. It does not guarantee agreement with real sources.
How does LiveCanvas differ from LiveChart?
It is the same project after a rename: LiveChart made charts, and LiveCanvas also works with images, emoji, scenes and processes. The old `$livechart` call is preserved if you put the livechart folder alongside.
How long is the animation and what size?
By default 1080×1440 pixels, 30 frames per second and about three seconds without sound. These are starting values that can change to suit the card.
Discussion
No questions or reviews yet.
Related
SkillsA Claude Code skill for short animated videos drawn in code: seven styles, story and storyboard sign-off, voice-over and an MP4 render
A skill that turns Chinese story copy or a set of ordered images into a vertical hand-drawn animation, a silent diary-comic picture track
A skill for Claude Code and Codex that builds cinematic product videos with Remotion from shot recipe cards, a ready template and audio assets
A Codex skill that redraws your photo as a watercolor illustration on white paper, with wide margins and a handwritten caption
A skill and Remotion pipeline: the agent studies your product and builds a demo video in React code, with no screen recording
A skill that turns a topic into a narrated 4K video: research, script, local TTS, Remotion composition and an FFmpeg mix
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| LiveCanvas (this entry) | no | 539 | |
| Claude Remotion Skill | no | 696 | teaches an agent to make video in Remotion in general, without fact search or a Live Photo package |
| Animate | no | 1,066 | draws short clips in code in several styles, without data verification or iPhone assets |
README in short
The description shows fifteen sample animations as GIFs and explains how to download the MP4 files and Live Photo packages. Installation, a sample request, environment requirements and the contents of the result follow. It ends with the MIT license for the code, terms for the sample media and credits.
README badge
Original SKILL.md
--- name: livecanvas description: 将主题或材料经搜索核实制成苹果 Live Photo 信息动效卡。用 Remotion 编排图片、PNG、emoji、场景、时间线和可选图表，适用于知识解释、产品比较、变化过程及数据故事；制作可预览动效与配对实况资源，说明接收端兼容性。 --- # LiveCanvas · 实况画布 把用户的一句话变成有证据、静态可读、动态能讲清内容的实况动效卡。原名 LiveChart，现名 LiveCanvas（实况画布），覆盖图表、图片与场景叙事；默认主动设计与主题相关的图片/图标和场景动作，图表只是可选表达。默认中文、Remotion、1080×1440、30 fps、约 3 秒、无声。这些是起点，允许按内容调整；3 秒不是声称的苹果格式限制。 ## 先定内容，再定动效 1. 从请求提取主体、指标、时间范围、比较关系、用途；给定文章或报告时先读完整材料，再提取关键事件、变化、能力、对比与场景。按内容重要性分配篇幅，不只挑价格或易展示的数字。具体需求优先，不把用户给的例子当成固定模板。 2. 搜索并核实数据，读 [数据与口径](references/data.md)。先保存原始证据、事实或结构化数据，再写标题、数字与注释。无时间序列的主题可以用事实卡、过程或场景解释；不能为了画图虚构数据。 3. 写简短制作说明：整套内容主线、每张的主体与结论、事实/数据口径、封面信息层级、分镜时间表及素材计划。张数依用户要求和内容决定，不固定为某个数量或主题组合。选出适合该主题的表达，不默认折线图加数字。直接做设计判断；仅当缺失信息会实质改变答案时提问。 4. 按 [封面与动效](references/design.md) 制作。吸收 Emil Kowalski animate 的有目的、克制、连续、节奏清晰原则；用户已指定 Remotion，不执行上游的工具替换步骤，也不把交互组件时长限制机械套到解释性视频。 ## 文案：直接写事实与数据 - 标题优先写“主体 + 数字/变化/明确结论”，例如“输出上限 1M tokens”“首发任务成本低 39%”“Text Arena 暂列第一”。没有必要的数字时，直接写产品名、版本或具体事件，不硬凑数据。 - 用简短、直观的陈述，删掉空泛宣传、设问、比喻、对仗和装饰性英文标签。不要用“重新定义”“读懂三个变化”“让收益落到真实场景”等句子代替事实。正文与配图描述也遵循同一标准；来源写清机构名，如 Artificial Analysis，不只写 AA。写法示例与限定条件见 [封面与动效](references/design.md#标题与配图文案)。 - 用户说“AI 味太浓”“直接上数据”“不要花里胡哨”时，优先精简文案与无关装饰；保留原有主题、信息、必要素材和已选设计方向，不擅自删除主体封面、合并主题或把整套改成同一数据模板。 - 修改时保留已确认的主题、信息与视觉要求。用户改变张数或重点时重新分配内容，并说明合并、替换与复用关系；不靠机械拆页或重复主题凑数。按图编号核对改版前后内容，总包按不同内容去重。 ## 视觉表达 - 封面以产品名、版本号或核心对象建立视觉主次。“登场感”“冲击力”等是效果描述，不直接当成画面文字；具体做法见 [封面与动效](references/design.md)。 - 整套统一配色、字体和图形语言，按内容选择构图。用图形解释事实与关系，避免堆文字、大数字、装饰卡片或机械复用同一模板。 - 默认让主题中的对象出现在画面中：产品 PNG、咖啡杯、汽车、列车、奖牌、城市照片、天气 emoji 等；挑选与内容有关的一组统一风格素材，不是给图表角落贴装饰。 - 设计对象的入场、运动、排列、替换或状态变化来解释内容。例如列车驶过年份节点、咖啡杯带出门店卡片、城市画面随季节切换。 - 按主题组合场景、图片对比、图标阵列、时间线、过程演示、重点文字和必要图表。纯图表只在确实最清楚或用户指定时采用。 - 使用图片或 emoji 时读 [图片与图标](references/visual-assets.md)，负责找素材、检查可用性并加入工程；无需让用户先提供素材才开始。 ## 搜索默认值 - “最近 3 年”：截至最新可靠已完成观测日的滚动三年，注明首尾实际日期，不默认成三个年末值。月度图用已完成月末观测；最新未完成月只能单独注明。 - “全年温度”未给年份：采用最近一个完整自然年，默认月平均气温；在结果开头说明选择。若意图明显是常年气候，使用有明确基准期的气候平均值。 - 跨公司比较对齐观测日期、指标、币种，或明确使用同基期指数，保留原始绝对值。 - 允许涨、跌、持平、季节循环。不要因用户说“增长曲线”而把下降画成增长。 - 日期按用户时区解析，不用机器日期硬编码“今年”。 ## 制作与封装 读 [运行与交付](references/production.md) 和 [直接交付](references/direct-delivery.md)。资源路径相对于本 skill 目录；称该绝对路径为 `SKILL_DIR`。