cyberneticlibrary
← library

Make calls, transcribe, and create voiceovers

ops-voiceskillsetup L2★3,035
davepoon/buildwithclaude ↗
What it does

Execute phone calls and voice transcription

Best for

Voice-enabled operations replacing legacy OpenClaw with Bland AI calls, ElevenLabs TTS, and live transcription.

Inputs
  • · Sub-command: call|tts|transcribe|setup
  • · Phone number or audio file path
  • · Prompt text or audio stream
Outputs
  • · Call recording and transcript
  • · TTS audio file with ElevenLabs voice ID
  • · Transcribed text from Whisper/Groq
Requires
  • · Bland AI API
  • · ElevenLabs API
  • · Whisper/Groq
Preconditions
  • · Bland AI API key, ElevenLabs key, or Whisper server running
  • · Credential resolution: userConfig → env → Doppler → password manager
  • · Audio I/O available for recording/playback
Failure modes
  • · Missing API key produces Doppler CLI fallback with latency
  • · Bland AI call timeout if phone unreachable
  • · Whisper quota exceeded blocks transcription mid-stream
Trust signals
  • · Direct API integration with credential resolution fallback
  • · Test-first workflow with TDD phases documented
  • · Structured JSON output format