| name | telegram-voice-transcription |
| description | Transcribe saved Telegram voice notes and other audio attachments with OpenAI before replying. Use when a Telegram prompt includes an audio file path under inbox/telegram/. |
| compatibility | Requires OPENAI_API_KEY or HERMIT_OPENAI_API_KEY, network access to api.openai.com, and a saved local audio file. |
Telegram Voice Transcription
Use this skill when a Telegram message includes a saved audio attachment such as:
inbox/telegram/...voice-note.ogg
inbox/telegram/...audio.mp3
inbox/telegram/...video-note.mp4
What it does
- sends the local audio file to OpenAI transcription
- defaults to
whisper-1
- writes a sidecar transcript file next to the audio
- prints the transcript to stdout so you can use it immediately
Usage
From the repo root:
node skills/telegram-voice-transcription/transcribe.mjs workspace/inbox/telegram/<file>
Examples:
node skills/telegram-voice-transcription/transcribe.mjs workspace/inbox/telegram/00000011-000013-voice-note.ogg
node skills/telegram-voice-transcription/transcribe.mjs workspace/inbox/telegram/00000011-000013-voice-note.ogg --model whisper-1
node skills/telegram-voice-transcription/transcribe.mjs workspace/inbox/telegram/00000011-000013-voice-note.ogg --stdout-only
Output
By default the script writes:
<audio-file>.transcript.md
The file contains:
- source audio path
- model used
- timestamp
- transcript text
Recommended workflow
- Run the script on the saved Telegram audio file.
- Read the transcript.
- Answer the human based on the transcript.
- Use
send_telegram_message or send_telegram_voice_note for the reply.
Notes
- If the audio is unclear, say that plainly instead of pretending it was clear.
- If transcription fails because the API key or network is missing, tell the human and ask for text instead.