voice-handling

Transcribe on-device audio to text using a local Whisper pipeline.

Updated Mar 9, 2026

Search Tags:#voice #transcription #whisper #privacy #on-device #telegram

Ready to run

npx skills add https://github.com/dbbaskette/herald --skill voice-handling

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: voice-handling
Source: https://github.com/dbbaskette/herald/tree/main/skills/voice-handling
Command: npx skills add https://github.com/dbbaskette/herald --skill voice-handling

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Transcribe voice memos and short audio clips to text locally, preserving privacy by keeping data on-device.

Core Features & Use Cases

On-device transcription: transcribes voice memos and short audio clips without sending data to external services.
Codec & format handling: converts common formats via ffmpeg when needed.
Use Case: when a user sends a voice message in Telegram, Herald transcribes it and makes the text available for further reasoning.

Quick Start

Transcribe the latest voice memo by selecting the audio file and running the transcription workflow.

Frequently Asked Questions about voice-handling

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

FAQPage Schema

How do I transcribe Telegram voice messages locally without uploading audio?▼

You can transcribe Telegram voice messages locally using an on-device Whisper pipeline. This processes audio files directly on your hardware, converting voice to text without sending data to external services.

What's the best way to ensure privacy when converting voice memos to text?▼

On-device transcription ensures privacy by keeping audio data on your local machine. Using a local Whisper setup processes voice memos and short audio clips without relying on cloud-based transcription services.

Do I need ffmpeg to transcribe short audio clips with a local Whisper setup?▼

You need ffmpeg to handle common codecs and formats when transcribing short audio clips. It converts audio files as needed before the local Whisper pipeline processes them into text.

Can I use whisper.cpp for on-device voice transcription?▼

Yes, you can use whisper.cpp for on-device voice transcription. The skill supports local processing using either openai-whisper or whisper.cpp to transcribe audio files while protecting privacy.

Does on-device transcription support common audio codecs and formats?▼

On-device transcription supports common audio codecs and formats by utilizing ffmpeg for conversion when needed. This ensures various short audio files are processed correctly by the local Whisper pipeline.

What are the limitations of local Whisper voice transcription?▼

Local Whisper voice transcription requires an on-device setup with openai-whisper or whisper.cpp and optional ffmpeg. It is designed for voice memos and short audio clips, enforcing local processing without uploading audio to external servers.