skills.rest
cosmicstack-labs
cosmicstack-labsOFFICIAL
Explore all skills by this author →
Website𝕏@cosmicstackGitHub

hyperframes-media

Generate local TTS audio, word-timestamp transcripts, and transparent overlays for media compositions.

364|53|Updated May 9, 2026
Search Tags:#text-to-speech#captions#transcription#whisper#background-removal#hyperframes#media-preprocessing
One-click install
npx skills add https://github.com/cosmicstack-labs/mercury-agent-skills --skill hyperframes-media-cosmicstack-labs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hyperframes-media
Source: https://github.com/cosmicstack-labs/mercury-agent-skills/tree/main/categories/development/hyperframes-media
Command: npx skills add https://github.com/cosmicstack-labs/mercury-agent-skills --skill hyperframes-media-cosmicstack-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill removes the manual effort of preparing media assets by generating local text-to-speech audio, transcribing it into timestamped text, and removing backgrounds for transparent overlays.

Core Features & Use Cases

  • Local text-to-speech (TTS) narration: Creates voiceover audio from text using Kokoro-82M without requiring an API key, with support for voice selection and language/locale-aware behavior.
  • Word-level transcription for captions: Converts audio/video or subtitle files into normalized transcript JSON with precise timestamps, including guidance to avoid accidental language translation.
  • Background removal for transparent media: Produces cutouts and transparent video layers (e.g., VP9 alpha WebM or ProRes 4444) using u2net, suitable for composing titles/graphics over subjects.

Use Case: Generate a voiceover from a script, transcribe the narration into word timestamps, and then create caption-ready timing and transparent overlays for a HyperFrames composition.

Quick Start

Ask the AI to generate narration audio from your script, transcribe the resulting audio into word-level timestamps, and output it for captions and composition use.

Frequently Asked Questions about hyperframes-media

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate voiceover audio from a script locally without an API key?▼

You can generate local text-to-speech voiceover audio from a script without an API key using the Kokoro-82M model. The Skill supports configurable voice selection and language-aware behavior, downloading and caching the required TTS model on first run.

How do I transcribe audio into word-level timestamps for captions?▼

To transcribe audio into word-level timestamps for captions, the Skill uses Whisper models to convert audio, video, or subtitle files into normalized transcript JSON with precise timing. It includes guidance to prevent accidental language translation during transcription.

Can I remove backgrounds to create transparent video overlays for composition?▼

Yes, you can remove backgrounds to create transparent video overlays using the u2net model. The Skill produces cutouts and alpha-capable output formats like VP9 alpha WebM or ProRes 4444, suitable for layering titles or graphics over subjects.

Do I need to install TTS, Whisper, or u2net models before generating media assets?▼

No, you do not need to pre-install TTS, Whisper, or u2net models before generating media assets. The Skill automatically downloads and caches the required models on first run, supporting configurable transcription model selection.

What is the best way to automate narration, transcription, and cutout creation for a composition?▼

The best way to automate narration, transcription, and cutout creation is to process scripts sequentially: generate TTS audio, extract word-timestamp transcripts, and create background-removed overlays. This produces caption-ready timing and transparent layers for composition workflows.

Related Skills

Explore More Design & Creative Skills→
SpaceZephyrSpaceZephyr
333

ffmpeg-usage

Convert, edit, compress, and transcribe media files with FFmpeg.

Community
Advanced
compozycompozy
182

hyperframes-media

Generate TTS audio, word-timestamp transcripts, and transparent overlays for HyperFrames.

Official
Advanced
JosiahSiegelJosiahSiegel
51

ffmpeg-animation-timing-reference

Reference FFmpeg and ASS/SSA animation timing units and durations.

Community
Intermediate
hoanghd218hoanghd218
7

hyperframes-media

Generate local voiceovers, transcripts, and background-removed overlays with CLI commands.

Community
Intermediate