GuideSlate Editorial Library — Full English AI tool tutorials are 100% free · Companion .zip starter kits $9.90 one-time 全部工具包
ComfyUI & video Updated Nov 2026 · v1.9 · Saves 8 hrs · Intermediate

Render batch product demo videos from JSON with Claude Code and Remotion

Replace slow timeline editing and generic templates: mount a Remotion Agent Skill in Claude Code, feed a JSON payload of feature highlights, and render crisp 60fps MP4 clips from the terminal.

ARCHITECTURE PIPELINE
01 · SpecFeature JSON02 · AgentRemotion SKILL.md03 · Spring60fps React frame04 · RenderBatch H.264 MP4

Why default setups fall short

One of the most bookmarked creator workflows on X combines Anthropic's Claude Code with Remotion's programmatic React video engine. Instead of dragging keyframes inside After Effects or CapCut, you lock physics-based spring curves inside a reusable SKILL.md and let a Node worker render 10 to 100 localized product launch videos directly from a JSON array.

Prerequisites before starting

  • Node.js 20+ and FFmpeg installed locally or in Docker
  • Claude Code, Cursor, or Antigravity for component generation
  • A JSON file containing product titles, metrics, and UI screenshots

Step-by-step implementation

STEP 01

1. Lock spring physics and forbid CSS transitions in SKILL.md

Standard CSS animations drop frames during headless Chromium video capture because they run on wall-clock time instead of frame index. Your Remotion skill file must force the coding agent to drive every opacity, scale, and translateY value strictly through useCurrentFrame() and spring().

// Hard rule inside 01-remotion-motion-SKILL.md
const frame = useCurrentFrame();
const { fps } = useVideoConfig();
const entrance = spring({
  frame: frame - 12,
  fps,
  config: { damping: 18, stiffness: 140, mass: 0.8 }
});
STEP 02

2. Parameterize compositions with Zod inputProps schemas

Define a strict schema for your video composition so every text headline, badge metric, and accent hex can be overridden at render time via CLI flags or a JSON batch queue without touching React code.

export const DemoCardSchema = z.object({
  headline: z.string(),
  metricValue: z.string(),
  subcopy: z.string(),
  accentColor: z.string().default("#1B4332")
});
STEP 03

3. Run headless batch rendering across a JSON manifest

Use a lightweight Node runner script that loops over your JSON array and invokes npx remotion render with --props, outputting social-ready 1080x1350 or 1920x1080 H.264 MP4 files with constant rate factor (CRF 18) for zero compression artifacts on X.

npx remotion render src/index.ts ProductLaunchCard out/launch-01.mp4 \
  --props='{"headline":"Shipped in 48 hours","metricValue":"+310%","subcopy":"Zero timeline editing"}' \
  --codec=h264 --crf=18
STEP 04

4. Verify audio loudness and first-frame thumbnail hook

X auto-plays video muted in the feed. Always render frame 0 with the primary metric card already at 85% opacity so the static feed thumbnail is readable before playback starts.

More practical field guides

View all 12 guides
Claude & Cursor skills Featured kit
Saves 4.5 hrs · v2.4
Replace generic AI interfaces with a custom SKILL.md design token system

How top solo builders force Claude Code, Cursor, and Antigravity to output studio-grade editorial interfaces instead of default blue-gray Tailwind cards.

01 · InputSKILL.md rules02 · Tokens6-hex palette03 · GuardrailAnti-cliche lint04 · OutputEditorial SSR UI
Companion .zip starter kit (4 files) $9.90 one-time
01-editorial-frontend-SKILL.md (6.4 KB)02-cursor-anti-slop-rules.mdc (3.8 KB)03-architectural-bento-tokens.css (5.1 KB)README-quickstart.md (2.2 KB)
Read free guide
n8n & automation
Saves 6 hrs · v3.1
Monitor X and Reddit buying-intent threads every 15 minutes with n8n

Skip $99/mo social listening SaaS tools. Poll targeted RSS and search endpoints, score high-intent leads with a low-cost Flash LLM, and push ready-to-reply alerts to Slack.

01 · Cron 15mSearch feeds02 · DedupSQLite hash03 · Flash LLMScore >= 8/1004 · WebhookSlack alert
Companion .zip starter kit (4 files) $9.90 one-time
01-n8n-intent-radar-v3.json (11.2 KB)02-intent-classifier-prompt.md (3.4 KB)03-docker-compose.n8n.yml (1.9 KB)README-quickstart.md (2.0 KB)
Read free guide