技能详情（站内镜像，无评论）
作者：Vega @babysor
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 0 · 231 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：可疑 · OpenClaw ：可疑
Package：babysor/speak1
安全扫描（ClawHub）
- VirusTotal ：可疑
- OpenClaw ：可疑
OpenClaw 评估
The skill's purpose (local TTS) matches its instructions, but the runtime instructions require downloading and running binaries (and an installer) not declared in the manifest, so the package metadata and runtime behavior are inconsistent and merit caution.
目的
Name/description (local TTS via Kokoro) match the instructions: examples and reference describe a CLI kokoro-tts and model files needed to convert text/epub/pdf to audio.
说明范围
SKILL.md tells the agent to install kokoro-tts (via 'uv tool install kokoro-tts') and to wget model files from GitHub releases into the working directory. The instructions reference using kokoro-tts, wget, and Python (3.9–3.12) at runtime, but the skill metadata does not declare these requirements. Downloads and saving model binaries to disk are explicitly required and could install/execute third-party code.
安装机制
No install spec in registry metadata, but the instructions call out 'uv tool install kokoro-tts' and direct downloads from GitHub release URLs (github.com/nazdridoy/kokoro-tts/releases/...). GitHub releases are a reasonable host, but 'uv tool install' is an unspecified installer; without knowing what that installer does, there's moderate risk. The downloads are direct model/binary assets (onnx, .bin) — no archive extraction specified.
证书
The skill requests no environment variables or credentials, which is appropriate for local TTS. However metadata omission of required runtime tools (kokoro-tts, wget, python) is an inconsistency to address.
持久
always:false and no declared installs that modify other skills or system-wide config. The skill instructs placing model files in the working directory (normal for local models) but does not request persistent elevated privileges.
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「speaker-local」。简介：Text-to-speech using Kokoro local TTS. Use when the user wants to convert text …。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/babysor/speak1/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
---
name: speak
description: Text-to-speech using Kokoro local TTS. Use when the user wants to convert text to audio, read aloud, or generate speech.
---
# speak
Convert text to speech locally with [Kokoro TTS](https://github.com/nazdridoy/kokoro-tts).
## Quick start
```bash
# Text string → audio file
kokoro-tts <(echo "Hello world") hello.wav --voice af_sarah
# Text file → audio
kokoro-tts article.txt out.wav --voice af_heart
# Chinese
kokoro-tts story.txt out.wav --voice zf_xiaoni --lang cmn
# EPUB / PDF → chapter audio files
kokoro-tts book.epub --split-output ./chapters/ --format mp3 --voice bf_emma
# Voice blending (60-40 mix)
kokoro-tts input.txt out.wav --voice "af_sarah:60,am_adam:40"
# Adjust speed
kokoro-tts input.txt out.wav --voice am_adam --speed 1.2
# Stream playback (no file saved)
kokoro-tts input.txt --stream --voice af_nova
```
## Install
```bash
uv tool install kokoro-tts
```
Model files (`kokoro-v1.0.onnx`, `voices-v1.0.bin`) must be in the working directory. Download once:
```bash
wget https://github.com/nazdridoy/kokoro-tts/releases/download/v1.0.0/kokoro-v1.0.onnx
wget https://github.com/nazdridoy/kokoro-tts/releases/download/v1.0.0/voices-v1.0.bin
```
## Voices at a glance
| Region | Female | Male |
|--------|--------|------|
| 🇺🇸 en-us | af_alloy af_heart af_sarah af_nova ... | am_adam am_echo am_michael ... |
| 🇬🇧 en-gb | bf_alice bf_emma bf_lily ... | bm_daniel bm_george ... |
| 🇨🇳 cmn | zf_xiaoni zf_xiaoxiao zf_xiaoyi ... | zm_yunxi zm_yunyang ... |
| 🇯🇵 ja | jf_alpha jf_nezumi ... | jm_kumo |
| 🇫🇷 fr-fr | ff_siwis | — |
| 🇮🇹 it | if_sara | im_nicola |
For full voice list, options, and input format details, see [reference.md](reference.md).