Francisco-Andre-Martins/Scriberr ? reverse-engineered prompt

Build me a self hosted web app for private audio transcription that runs locally and keeps everything off the cloud.

I want to be able to upload a recording or paste a YouTube link, get a clean transcript back, and read it in a simple interface with audio playback that follows along with the text. It should also support speaker labels, word timing, highlights, and quick notes that jump back to the right spot in the transcript and audio.

Please add a way to summarize transcripts and chat with them using either OpenAI or a local model like Ollama. I’d also like reusable transcription profiles, batch uploads, a quick transcribe mode for one off files, and downloads as text, SRT, JSON, and similar formats.

Include a REST API with API key management, and make the app easy to run in Docker or as a single binary on my own machine. If you need current docs for transcription or diarization tools, look them up online first.