skills.rest
crisandrews
crisandrewsCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

voice

Configures and tests text-to-speech and transcription backends for an agent.

62|14|Updated Apr 9, 2026
Search Tags:#text-to-speech#transcription#whisper#audio#elevenlabs#openai tts#voice setup
One-click install
npx skills add https://github.com/crisandrews/ClawCode --skill voice-crisandrews
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: voice
Source: https://github.com/crisandrews/ClawCode/tree/main/skills/voice
Command: npx skills add https://github.com/crisandrews/ClawCode --skill voice-crisandrews

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Setting up voice capabilities for an AI agent involves choosing TTS/STT backends, managing API keys safely, and avoiding duplicate transcription work across messaging channels. This Skill guides the agent through diagnosing, configuring, and testing voice so users don't have to figure out backend wiring themselves. ## Core Features & Use Cases - Status diagnostics: Run /agent:voice status to see which TTS/STT backends (sag/ElevenLabs, OpenAI TTS, macOS say, whisper-cli) are available and whether voice is enabled. - Guided setup: Step-by-step instructions for installing backends, setting environment variables like ELEVENLABS_API_KEY or OPENAI_API_KEY, and enabling voice.enabled in config. - Voice testing: Generate a sample spoken greeting via voice_speak and report the output file path, optionally sending it through a messaging channel. - Use Case: A user says "configurar voz" and the agent checks status, recommends sag via Homebrew, walks them through adding their ElevenLabs key to their shell rc, then runs a test greeting in Spanish. ## Quick Start Ask the agent to run /agent:voice status to check which voice backends are available and whether voice is enabled.

Frequently Asked Questions about voice

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

FAQPage Schema
How do I set up text-to-speech for my AI agent?▼

Run /agent:voice setup and the agent checks available backends, then recommends sag (a Homebrew wrapper around ElevenLabs), OpenAI TTS via OPENAI_API_KEY, or the built-in macOS say command as a zero-setup fallback.

How do I test if agent voice output is working?▼

Run /agent:voice test. The agent confirms voice is enabled and a backend is available, calls voice_speak with a greeting in your language, and reports the generated audio file path.

Should I use voice_transcribe on WhatsApp voice notes?▼

No, if the WhatsApp plugin has audio transcription enabled it is authoritative for that channel. Calling voice_transcribe there duplicates work and produces two different transcriptions. Use it for WebChat uploads, iMessage audio, or standalone files.

Where should I store ElevenLabs or OpenAI API keys?▼

Store them as environment variables in your shell rc file (~/.zshrc or ~/.bashrc), never in agent-config.json, since that config file may end up in a git repository. Restart the agent after adding them.

What speech-to-text options work offline?▼

whisper-cli, installed via brew install whisper-cpp, provides free offline transcription. The alternative is the OpenAI Whisper API, which requires an OPENAI_API_KEY and network access.

Related Skills

Explore More Software Engineering Skills→
trpc-grouptrpc-group
1.8k

sag

Generates ElevenLabs text-to-speech audio with mac-style say command UX.

Official
Basic
adamhjoujadamhjouj
4

bluebubbles

Automate sending and managing iMessages via BlueBubbles through OpenClaw's message tool.

Community
Intermediate
sliamh11sliamh11
50

use-local-whisper

Switch voice transcription from OpenAI's Whisper API to local whisper.cpp.

Community
Intermediate
erguotou520erguotou520

add-voice-transcription

Transcribe WhatsApp voice notes to text using OpenAI Whisper.

Community
Advanced