HariNayan

Script To Reel — AI skill for Claude Code

AI community

Turn a written script into a finished 9:16 vertical reel.

How to install Script To Reel

This entry records only its repository, not the path inside it, so there is no exact command to give. Open HariNayan/Script-to-Reel and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Script To Reel does

Turn a written script into a finished 9:16 vertical reel. An AI-powered Claude skill that generates an illustrated cartoon host, per-beat scene art, labelled props, and word-timed burst captions from a voiceover. Built as a HyperFrames composition and rendered to MP4. Supports Hindi, English, and code-switched speech.

Alternatives in AI

  • Hindsight — State-of-the-art long-term memory for AI agents by Vectorize 6.7k ★
  • Pixel Agents — Compressed Reference — VS Code extension with embedded React webview: pixel art office where AI agents (Claude Code terminals) are an 6.5k ★
  • Rb Setup — First-time setup 1.3k ★

README

Script to Reel

Turn a written script into a finished 9:16 vertical reel — an illustrated cartoon host, per-beat scene art, labelled props, and word-timed burst captions — built as a HyperFrames composition and rendered to MP4.

This is a **Claude skill** that automates the pipeline from script to short-form video, designed for Reels, Shorts, TikTok, and vertical explainers.

What It Does

Takes a script and produces a vertical narrated explainer:

  • A flat-vector cartoon host who gestures through the beats
  • One illustrated scene per beat with a text label
  • Burst captions (2-3 words) locked to the real voiceover timing
  • Everything rendered from authored SVG — resolution-independent, no design drift

Pipeline Overview

Script → Audio → Word Onsets → Caption Chunks → SVG Art + Animation → MP4
Step Command What Happens
1. Timing python scripts/word_timings.py Extract word-level onsets via Whisper
2. Captions python scripts/build_captions.py --cues cues.json --onsets onsets.json Chunk into 2-3 word burst frames
3. Art Author SVG symbols Character rig + scene props
4. Motion GSAP timeline Pose swaps, prop entrances, caption pops
5. Verify npm run check && npx hyperframes snapshot Validate and preview

Scripts

`word_timings.py`

Extracts word-level onset timestamps from a voiceover file using OpenAI Whisper.

python scripts/word_timings.py assets/voiceover.mp3 --language hi --out onsets.json
  • Supports Hindi, English, and code-switched speech
  • Transcript text is deliberately discarded — only onset times are used (ASR garbles code-switched speech)
  • Outputs onsets.json with word timestamps as a timing grid

`build_captions.py`

Chunks sentence-level caption cues into 2-3 word burst frames, snapping boundaries to real speech onsets.

python scripts/build_captions.py --cues cues.json --onsets onsets.json
`