# Ppt Master Data Journalism Skill

> PPT Master Data Journalism Skill

- Skill: `chaitin/ppt-master-data-journalism-skill` (Agent Skill, multi-file: 6 files)
- Install (CLI): `npx skillmds@latest add chaitin/ppt-master-data-journalism-skill`
- Raw SKILL.md: https://api.skillmd.com/api/skills/chaitin/ppt-master-data-journalism-skill/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: chaitin (https://skillmd.com/u/chaitin)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/chaitin/ppt-master-data-journalism-skill

---

# PPT Master Data Journalism Skill

## Scope

This built-in skill package owns the `data-journalism` template variant imported from PPT-Master.

- `skill_id`: `ppt-master-data-journalism-skill`
- `variant_id`: `data-journalism`
- upstream example: `/Users/mawenyi/try/ppt-master/examples/ppt169_global_ai_capital_2026`
- upstream final SVGs: `svg_final/*.svg`
- upstream execution lock: `spec_lock.md`
- package renderer: `assets/template.html`
- package validator: `scripts/validate-deck.mjs`
- layout skeletons: `references/layouts.md`
- QA checklist: `references/checklist.md`

## Design Contract

Use this package for Data Journalism / 数据新闻 decks. The source template is the 20-page *2026 全球 AI 资本格局* industry briefing, not a light generic dashboard theme. Slides should preserve the PPT-Master fixed 1280x720 canvas model and use `.pm-canvas.pm-data-news` as the visual root.

Required visual language, derived from the upstream `design_spec.md` and `spec_lock.md`:

- 深色 Bloomberg/Economist 新闻信息图：`#0E1116` graphite background, `#1A1F26` panels, `#E8E6E1` primary text, `#E63946` finance red, `#F4A261` amber secondary accent.
- Cambria-like editorial serif titles and hero numbers; Microsoft YaHei / Arial dense body copy; Consolas-like source lines and chart labels.
- Header area uses section label + thin divider; footer carries page number and source notes. The deck reads like a night-reading financial longform, not a keynote dashboard.
- Use the original 20-page rhythm: cover, editor note, contents, Q1 VC KPI landscape, hyperscaler capex bar chart, big-three valuation dumbbell, valuation-vs-ARR bubble, OpenAI investor donut, big-three comparison, China five ranking, US/China dual track table, second-tier cards, Nvidia chapter, Nvidia sankey, Nvidia Pareto, Stargate hub, bubble chapter, capex-vs-revenue dual-axis, four-risk quadrant, six-takeaway close.
- Every data page must include a source or methodology line and one package-local evidence component such as `.pm-data-chart`, `.pm-data-card`, `.pm-data-bar`, `.pm-data-rank`, `.pm-data-metric`, `.pm-data-source`, `.pm-data-method`, `.pm-data-finding`, `.pm-data-axis`, `.pm-data-callout`, `.pm-data-table`, `.pm-data-node`, or `.pm-data-flow`.
- Data must stay source-grounded. For the original Global AI Capital example, use the locked facts from the upstream outline: `$297B`, `81%`, `$188B`, `65%`, `$725B`, `+77%`, `$122B`, `$852B`, `$40B+`, `85%`, `$500B`, `10GW`, `+46pt`, and the named sources Crunchbase, Bloomberg, CNBC, TechCrunch, Anthropic, and OpenAI.
- Motion only uses declarative hooks: `data-pm-motion="data"` and `data-pm-anim`.

Do not generate unsourced exact values, generic business card grids, light beige report pages, glass panels, Memphis stickers, dark-tech terminals, animation JavaScript, `@keyframes`, external scripts, or custom runtime code in generated slide fragments.