How is authentication handled for the API?

It uses JWT (JSON Web Tokens) passed via the x-jwt-token header. Tokens can be stored locally in a config file or passed through environment variables.

Does it support identifying different speakers?

Yes, by enabling the diarization flag and specifying the maximum number of speakers, the skill can distinguish between different voices in a recording.

What audio formats can this skill transcribe?

The skill supports common formats like OGG and WAV. If a format is unsupported by the backend, it provides FFmpeg commands to convert files to compatible 16kHz WAV format.

Is this skill suitable for long recordings?

While defaults are optimized for short voice notes (under 5 minutes), the polling mechanism allows it to handle longer tasks asynchronously until the result is ready.

Speech2Text Client

Name: Speech2Text Client
Author: kvcop

bykvcop

0•

Data Science & ML

Transcribes audio and video files into text via a specialized Speech2Text API with support for JWT authentication and task polling.

The Speech2Text Client skill enables Claude to interact with transcription services to convert voice notes, video audio, and recordings into structured text. It handles the end-to-end transcription lifecycle, including user registration, secure JWT authentication, file uploads, and status polling for asynchronous tasks. While optimized for single-speaker voice notes—particularly those originating from Telegram—it also supports advanced features like multi-speaker diarization and provides troubleshooting guidance for media format conversions and service-level errors.

Key Features

01Integrated troubleshooting for media formats and server-side resource limits

02Support for multi-speaker diarization in complex meetings

03Asynchronous task polling and result retrieval for long recordings

04Secure JWT-based authentication and token management

050 GitHub stars

06Automated audio/video file upload and transcription task creation

Use Cases

01Automating the extraction of text from video content for documentation purposes

02Transcribing meeting recordings to generate summaries or action items

03Converting Telegram voice notes into searchable text for personal knowledge bases

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kvcop/ai-kb-assistant speech2text-client

For use in Claude.ai and ChatGPT

Download Skill