How do I install the Datamagic Video skill?

Run npx skillmds@latest add hkustdial/datamagic-video in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Datamagic Video safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Datamagic Video?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Datamagic Video free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Datamagic Video?

hkustdial (@hkustdial) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Datamagic Video by hkustdial — Agent Skill

Name: Datamagic Video
Author: hkustdial

Datamagic Video

Use when a task involves tabular data (CSV, Excel, database result, structured table) becoming a narrated data video, animated chart, data story, DVSpec, visualization storyboard, voiceover-timed chart sequence, Remotion data-video render, or refinement of an existing data-video result.

hkustdial 813f8af 18 files · 100.8 KB Updated Sep 17, 2026

File contents

DataMagic Video

Turn tabular data into a narrated, animated data video. Core artifact: DVSpec, a renderer-agnostic screenplay for scenes, data bindings, narration, and animation. Render with any stack; Remotion is reference path.

Use

CSV / Excel / structured table -> multi-scene narrated video
Pasted data -> animated chart or short data story
Existing DVSpec / render -> refine, debug, or improve
Goal only, no render requested -> produce DVSpec plan

Do not use for raw video editing, non-data text-to-video, or live dashboards.

Core Rules

Plan first. DVSpec before render.
Data leads story. Pick narrative pattern from data shape, not topic.
Narration leads viewer eye. Animation follows narration timing.
Read only needed rule files. Heavy details live under rules/.
Before calling work done: run rules/self-review.md.

Router

Task	Read
Profile data, find candidate insights	`rules/data-analysis.md`
Choose narrative pattern, scenes	`rules/scene-planning.md`
Pick chart type, axes, scales	`rules/chart-selection.md`
Author or validate DVSpec	`rules/dvspec.md`
Choose visual system, tokens, motion defaults	`rules/design-system.md`
Pick style pack	`rules/styles/README.md`
Render DVSpec to video with Remotion	`rules/remotion-integration.md`
Add TTS, audio-driven duration, subtitles	`rules/voiceover.md`
Write narration script	`rules/narration.md`
Refine existing output	`rules/refinement.md`
Diagnose bad output	`rules/anti-patterns.md`
Final quality gate	`rules/self-review.md`

Workflows

Full video:

data-analysis -> scene-planning -> chart-selection -> dvspec -> design-system -> styles -> remotion-integration -> voiceover -> narration -> self-review

DVSpec only:

data-analysis -> scene-planning -> chart-selection -> dvspec

Single animated chart:

data-analysis -> chart-selection -> dvspec -> design-system -> remotion-integration

Refinement:

refinement -> changed domain rule(s) -> self-review

Narrative Patterns

Pattern	Use when	Climax
`freytag_default`	One main finding builds over time	60-75%
`hook_then_evidence`	One striking number should open	first chart scene
`comparison_driven`	A vs B is story spine	largest contrast
`time_driven`	chronological change drives story	sharpest shift
`drill_down`	total -> segment -> cause	deepest zoom

See rules/scene-planning.md before locking pattern.

---name: datamagic-videodescription: Use when a task involves tabular data (CSV, Excel, database result, structured table) becoming a narrated data video, animated chart, data story, DVSpec, visualization storyboard, voiceover-timed chart sequence, Remotion data-video render, or refinement of an existing data-video result.---# DataMagic VideoTurn tabular data into a narrated, animated data video. Core artifact: **DVSpec**, a renderer-agnostic screenplay for scenes, data bindings, narration, and animation. Render with any stack; Remotion is reference path.## Use- CSV / Excel / structured table -> multi-scene narrated video- Pasted data -> animated chart or short data story- Existing DVSpec / render -> refine, debug, or improve- Goal only, no render requested -> produce DVSpec planDo not use for raw video editing, non-data text-to-video, or live dashboards.## Core Rules- Plan first. DVSpec before render.- Data leads story. Pick narrative pattern from data shape, not topic.- Narration leads viewer eye. Animation follows narration timing.- Read only needed rule files. Heavy details live under `rules/`.- Before calling work done: run `rules/self-review.md`.## Router| Task | Read ||---|---|| Profile data, find candidate insights | `rules/data-analysis.md` || Choose narrative pattern, scenes | `rules/scene-planning.md` || Pick chart type, axes, scales | `rules/chart-selection.md` || Author or validate DVSpec | `rules/dvspec.md` || Choose visual system, tokens, motion defaults | `rules/design-system.md` || Pick style pack | `rules/styles/README.md` || Render DVSpec to video with Remotion | `rules/remotion-integration.md` || Add TTS, audio-driven duration, subtitles | `rules/voiceover.md` || Write narration script | `rules/narration.md` || Refine existing output | `rules/refinement.md` || Diagnose bad output | `rules/anti-patterns.md` || Final quality gate | `rules/self-review.md` |## WorkflowsFull video:`data-analysis` -> `scene-planning` -> `chart-selection` -> `dvspec` -> `design-system` -> `styles` -> `remotion-integration` -> `voiceover` -> `narration` -> `self-review`DVSpec only:`data-analysis` -> `scene-planning` -> `chart-selection` -> `dvspec`Single animated chart:`data-analysis` -> `chart-selection` -> `dvspec` -> `design-system` -> `remotion-integration`Refinement:`refinement` -> changed domain rule(s) -> `self-review`## Narrative Patterns| Pattern | Use when | Climax ||---|---|---|| `freytag_default` | One main finding builds over time | 60-75% || `hook_then_evidence` | One striking number should open | first chart scene || `comparison_driven` | A vs B is story spine | largest contrast || `time_driven` | chronological change drives story | sharpest shift || `drill_down` | total -> segment -> cause | deepest zoom |See `rules/scene-planning.md` before locking pattern.

Frequently asked questions