Kokoro Agent Voices

基于 Kokoro TTS 的本地文本转语音 Skill，提供按 Agent 命名的音色映射，支持选择 54 种声音并输出 WAV 音频；首次运行需下载模型，之后本地推理，无 API 费用。

nissan

热度 58.67

最近更新 2026-05-18 12:11:04

分类：效率依赖：代码执行 / 读写文件 / 第三方 API应用场景：自动化脚本 / 文件转换 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add kokoro-agent-voices

Metadata

name: kokoro-agent-voices
version: 1.0.1
description: Local zero-cost text-to-speech with per-agent voice profiles using Kokoro TTS (82M params). 54 voices available, named agent mappings, WAV output. Use when building voice-enabled agents without API costs. Requires espeak-ng and a Python environment with kokoro dependencies.
metadata: {"openclaw":{"emoji":"🗣️","requires":{"bins":["python3","espeak-ng","espeak"],"env":[]},"primaryEnv":null,"network":{"outbound":true,"reason":"Downloads Kokoro model from Hugging Face Hub on first run (~350MB). All subsequent inference is local."},"security_notes":"Outbound network is used only once to download the Kokoro model weights from Hugging Face Hub (~350MB, one-time). All subsequent TTS inference runs entirely on the user's device. No audio data or text is ever transmitted externally."}}

Kokoro Agent Voices

Give each AI agent a distinct voice using Kokoro TTS — a lightweight 82M parameter model that runs locally with zero API costs. 54 voices across American, British, and other accents.

Agent Voice Profiles

AGENT_VOICES = {
    "loki": "am_fenrir",      # Deep, authoritative
    "archie": "bm_george",    # British analytical
    "sara": "af_bella",       # Warm, creative
    "kit": "am_echo",         # Clear, technical
    "liv": "af_nova",         # Bright, energetic
    "belle": "bf_emma",       # Refined, thoughtful
}

Usage

python3 scripts/speak.py --agent loki "System check complete"
python3 scripts/speak.py --voice af_bella "Hello world" --output /tmp/greeting.wav
python3 scripts/speak.py --list-voices    # Show all 54 voices
python3 scripts/speak.py --list-agents    # Show configured agent profiles

Setup

Requires a Python environment with kokoro, soundfile, and espeak-ng installed. The model downloads automatically from Hugging Face on first use (~350MB).

Files

scripts/speak.py — TTS script with agent profiles and voice selection