Browser APIs first
Speech recognition and speech synthesis use capabilities already available in modern browsers.
Code-powered creator tools
Useful tools for voice, video transcripts, and social content.
Start with practical utilities that run on browser APIs, deterministic templates, and lightweight parsing. No paid AI APIs, no hidden LLM calls.
1
Capture a voice note, paste a YouTube URL, or enter a raw idea.
2
Run local browser APIs and rule-based formatting.
3
Copy transcripts, posts, captions, and clip ideas.
Voice Tools
Speech to Text
Build principle
No LLM dependency in this starter repo.
Templates and heuristics
Social posts, repurposed content, and clip suggestions are generated with deterministic scoring and formatting rules.
Server only where needed
The YouTube transcript tool uses a small Node API to fetch public caption tracks because browsers cannot reliably do that directly.