메뉴
Renders each composed slide HTML to a single PDF page at 1920×1080, then merges them into deck.pdf. Uses headless Chromium via Bash. Validates the output has the right page count and resolution.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Brainstorm distinct video concepts for a brief — cinematic videos, brand films, motion accents, and background loops for product websites. Produces 3–5 genuinely different creative directions (not five flavors of one safe idea), each with a premise, the feeling it creates, why it fits the product and placement, and how hard it is to pull off in Higgsfield. Use this skill any time the work starts with "what could this video be" — concepting a hero background loop, a launch film, a section divider, an explainer, or a social spot — even when the user just says "I need a video for X." First stage of the Video Factory chain; hands the chosen concept to video-factory-storyboard. Pulls toward concrete subjects over the drone-over-city / glowing-AI clichés that AI video defaults to.
The main orchestrator for the Slalom Video Factory. Walks a user from "I need a video" to a ready-to-generate prompt package (and, on request, generated clips) via the Frame → Concept → Storyboard → Direct → Generate → Ship pipeline. Diagnoses intent, sequences the craft skills, and pauses for review at each handoff. Use this skill whenever someone wants a cinematic video, a brand film, a motion accent, or a background loop for a product website and hasn't already scoped it down to a single stage — "make me a hero video for X," "I need a background loop," "let's storyboard a launch film." For a single stage, route straight to that craft skill instead.
The motion taste layer for a video — lock one coherent look across every shot before any generation happens. Owns register, pacing, camera language, color and grade, light, grain/texture, and brand fit, so a multi-shot piece reads as one director shot it rather than a pile of clips. The motion analog of design-imagery. Use this skill any time a video needs art direction: setting the visual system for a brand film, keeping a background loop on-brand, grading decisions, or course- correcting clips that came back looking like different videos. Third stage of the Video Factory chain; reads a storyboard, writes a direction spec that video-factory-higgsfield-prompt turns into model parameters and prompt language. Actively rejects the drone-city / glowing-AI motion clichés.
Turn a directed storyboard into a ready-to-run Higgsfield prompt package and (when asked) execute the generations via the Higgsfield MCP. Owns model routing (seedance 2.0 / kling 3.0 default for video; nano banana pro / chatgpt image 2 for keyframes), the image-to-video prompt structure (motion, camera, atmosphere, sound over a start frame), parameter selection (resolution, duration, aspect, sound on/off, genre, end-frame for loops), negatives, retries, and web encoding handoff. Use this skill any time the task is "write the Higgsfield prompt," "generate this in Higgsfield," "what model and settings," "make the background loop," or turning shots into clips. Final craft stage of the Video Factory chain; reads concept + storyboard + direction and produces 40_Prompts.md plus generated keyframes and clips. Keyframe-first, loop- seam-aware, credit-disciplined.
Turn a chosen video concept into a shot-by-shot storyboard ready to generate in Higgsfield — shot list with timing, camera move, framing, transition, and a keyframe-still prompt for every shot (because Higgsfield video models are image-to-video and animate a start frame). Use this skill any time a video idea needs to become an executable shot plan: storyboarding a hero background loop, a brand film, an explainer, a section divider, or a social spot — even when the user says "break this into shots," "plan the shots," or "what's the shot list." Second stage of the Video Factory chain; reads a concept (or a one-line idea) and hands a storyboard to video-factory-creative-direction and video-factory-higgsfield-prompt. Designs loop seams for background video and protects headline safe-areas.
Senior naming expertise. Brainstorms brand names in three categories — literal, abstract, alliterative — with 15+ per category by default. Validates each candidate against existing brand collisions and industry-specific negative connotations (e.g., "Forge" for a law firm reads as "forged documents"). Screens for trademark and domain availability. Use when brainstorming, evaluating, or recommending names for products, companies, features, or offerings.
| name | presentation-factory-html-to-pdf |
| description | Renders each composed slide HTML to a single PDF page at 1920×1080, then merges them into deck.pdf. Uses headless Chromium via Bash. Validates the output has the right page count and resolution. |
| type | skill |
| plugin | presentation-factory |
| tags | ["type/skill","plugin/presentation-factory","scope/specialty","topic/pdf-export"] |
| status | active |
Called from the Ship stage orchestrator after the user has approved the composed slides.
Toolchain: ../../references/presentation-factory-toolchain.md. Foundations: ../../references/presentation-factory-foundations.md.
Bash check:
which chromium || which google-chrome || which chrome
which pdftk || which python3 # python3 + pypdf as fallback
If neither Chromium nor Chrome is found, surface the issue. macOS install:
brew install --cask chromium
For each slides/NN-template.html in the WIP folder:
chromium --headless --disable-gpu \
--no-margins \
--print-to-pdf="WIP/[deck]/_pdf/slide-NN.pdf" \
--virtual-time-budget=3000 \
--window-size=1920,1080 \
--hide-scrollbars \
--run-all-compositor-stages-before-draw \
"file://$PWD/WIP/[deck]/slides/NN-template.html"
Notes:
--virtual-time-budget=3000 waits 3 seconds for any JS (the chart template uses vanilla JS to render the SVG bars)--no-margins prevents Chrome's default print margins--window-size=1920,1080 matches the slide canvas exactly--hide-scrollbars removes any scrollbar artifactsChromium's default --print-to-pdf uses letter size. The HTML body is 1920×1080 px. With CSS @page { size: ... } in the template, we can force exact dimensions. The slalom.css tokens already include the 1920×1080 body sizing.
Add to the composed slide HTML at compose time (in the composer skill's pass — already covered in the composer), inside <style>:
@page {
size: 1920px 1080px;
margin: 0;
}
If pages still come out wrong, fall back to Puppeteer / Playwright via Node script:
// puppeteer-render.js
const puppeteer = require('puppeteer');
(async () => {
const browser = await puppeteer.launch();
const page = await browser.newPage();
await page.setViewport({ width: 1920, height: 1080, deviceScaleFactor: 2 });
await page.goto(`file://${slidePath}`, { waitUntil: 'networkidle0' });
await page.pdf({
path: outPath,
width: '1920px',
height: '1080px',
printBackground: true,
margin: { top: 0, right: 0, bottom: 0, left: 0 },
});
await browser.close();
})();
cd WIP/[deck]/_pdf
pdftk slide-*.pdf cat output ../deck.pdf
Fallback if pdftk not installed:
# merge.py
from pypdf import PdfWriter
writer = PdfWriter()
for f in sorted(os.listdir("WIP/[deck]/_pdf")):
writer.append(f"WIP/[deck]/_pdf/{f}")
writer.write("WIP/[deck]/deck.pdf")
# Count pages — must match slide count
pdfinfo WIP/[deck]/deck.pdf | grep "Pages:"
# Or via Python: PdfReader("deck.pdf").pages length
Also: visually spot-check the first slide by extracting its rendered preview:
pdftoppm -f 1 -l 1 -png -r 96 WIP/[deck]/deck.pdf WIP/[deck]/_preview
# Produces _preview-01.png — show to user as a sanity check
After the merged deck.pdf is in place, remove the per-slide intermediate PDFs in _pdf/ to keep WIP tidy (or leave them for debugging — config flag).
| File | Path |
|---|---|
| Merged deck | WIP/[deck]/deck.pdf |
| Preview thumbnail of slide 1 | WIP/[deck]/_preview-01.png (optional) |
| Render log | WIP/[deck]/render.log |
--print-to-pdf page dimensions defaulting correctly — always set @page in CSS and viewport in the headless commandslide-*.pdf glob sorts alphabetically; using NN-prefixed names ensures correct order