Feishu Edge Tts

使用微软 Edge TTS（免费）生成语音，发送到飞书。无需 API key，音质优秀，支持多语言多音色。

Security Scan

The skill's code and instructions largely match its stated purpose (generate Edge TTS audio and send to Feishu), but there is an inconsistency in the declared requirements/metadata vs. the actual runtime behavior: the scripts require Feishu credentials and network access even though the top-level registry metadata claimed no required env vars — the mismatch warrants caution.

Purpose & Capability

The name/description (Edge TTS → Feishu) match what the scripts do: use edge-tts, ffmpeg, and curl to generate audio and upload/send via Feishu APIs. Required tools (edge-tts, ffmpeg, curl) are appropriate for this purpose.

Instruction Scope

SKILL.md and scripts scope is limited to generating audio (edge-tts), converting to opus (ffmpeg), uploading to Feishu via open.feishu.cn, and sending a message. The runtime instructions do not read unrelated system files or attempt to exfiltrate data to unexpected endpoints.

Install Mechanism

No install spec; this is instruction + script based. Dependencies are installed via standard channels (pip for edge-tts, OS package managers for ffmpeg) which is low risk compared to downloading arbitrary binaries.

Credentials

The scripts require FEISHU_APP_ID, FEISHU_APP_SECRET, and FEISHU_CHAT_ID to upload/send audio—this is proportionate to sending messages via Feishu. However, the registry summary at the top incorrectly listed 'Required env vars: none' while clawhub.yaml and SKILL.md do declare/expect these env vars. That metadata mismatch is a red flag (could be oversight, but it affects informed consent before installation).

Persistence & Privilege

Skill does not request persistent elevated privileges, does not set always:true, and does not modify other skills or global configs. It runs as a normal script and cleans up temporary files.

What to consider before installing

This skill appears to do what it claims (generate Edge TTS audio and send it to Feishu). Before installing or running: - Note that you must provide FEISHU_APP_ID, FEISHU_APP_SECRET, and FEISHU_CHAT_ID as environment variables; the top-level registry metadata omitted these — don't trust that omission. - Create or use a Feishu app/account with the minimum permissions necessary (least privilege) for uploading and sending messages; avoid using highly privileged tenant credentials. - Inspect scripts yourself (or run them in an isolated environment) — the send_voice.sh script posts only to open.feishu.cn and uses local temp files, edge-tts, ffmpeg, curl, and python3 for JSON parsing. There are no hidden remote endpoints in the code. - Test with the --no-send flag first to confirm audio generation locally before providing real Feishu credentials. - Install edge-tts and ffmpeg from official package sources (pip and your OS package manager) to avoid malicious third-party binaries. If you want higher assurance, ask the author for canonical source (repo/homepage) and a signed release; the manifest claims a GitHub URL but 'Source' and 'Homepage' are unknown — lack of an authoritative upstream repo reduces confidence.

Feishu Edge TTS - 飞书语音条（微软免费 TTS）

使用微软 Edge TTS 生成语音，发送到飞书！完全免费！

🎯 功能特点

• ✅ 完全免费：使用微软 Edge TTS，无需 API key
• ✅ 音质优秀：微软 Azure 同款语音引擎
• ✅ 多音色支持：支持中文/英文/日文等多种语言
• ✅ 语音条格式：发送真正的飞书语音条（点击即播）
• ✅ 语速调节：支持 0.5x - 2.0x 语速
• ✅ 音调调节：支持音调高低调整

🎤 可用音色

中文音色

• zh-CN-XiaoxiaoNeural - 女声，温暖亲切（推荐）
• zh-CN-YunxiNeural - 男声，沉稳专业
• zh-CN-YunjianNeural - 男声，激情澎湃
• zh-CN-XiaoyiNeural - 女声，活泼可爱
• zh-CN-liaoning-XiaobeiNeural - 东北话女声
• zh-CN-shaanxi-XiaoniNeural - 陕西话女声

英文音色

• en-US-JennyNeural - 女声，美式英语（推荐）
• en-US-GuyNeural - 男声，美式英语
• en-GB-SoniaNeural - 女声，英式英语

更多音色

支持全球 100+ 语言，400+ 音色！

🔧 前置要求

1. Python 环境

# 安装 edge-tts
pip install edge-tts

2. Feishu 应用配置

同 Feishu Voice Skill

3. 系统依赖

# 安装 ffmpeg
yum install -y ffmpeg  # CentOS/OpenCloudOS
apt-get install -y ffmpeg  # Ubuntu/Debian

🚀 快速开始

步骤 1：安装依赖

pip install edge-tts

步骤 2：配置环境变量

export FEISHU_APP_ID="cli_xxx"
export FEISHU_APP_SECRET="xxx"
export FEISHU_CHAT_ID="oc_xxx"

步骤 3：发送语音

# 使用默认音色（女声）
bash scripts/send_voice.sh -t "主人晚上好～"

# 指定音色
bash scripts/send_voice.sh -t "Hello!" -v en-US-JennyNeural

# 调整语速
bash scripts/send_voice.sh -t "你好" --rate 1.2

# 调整音调
bash scripts/send_voice.sh -t "你好" --pitch 10

💡 使用示例

1. 温暖女声问候

bash scripts/send_voice.sh -t "主人早上好～ 新的一天开始啦，今天也要加油哦～" -v zh-CN-XiaoxiaoNeural

2. 专业男声播报

bash scripts/send_voice.sh -t "现在是北京时间上午 8 点，为您播报今日新闻。" -v zh-CN-YunxiNeural

3. 英文语音

bash scripts/send_voice.sh -t "Good morning! Have a nice day!" -v en-US-JennyNeural

4. 方言趣味

bash scripts/send_voice.sh -t "哎呀妈呀，这旮瘩真冷啊！" -v zh-CN-liaoning-XiaobeiNeural

📖 命令参数

bash scripts/send_voice.sh [选项]

选项:
  -t, --text <text>       要转换的文字（必需）
  -v, --voice <voice>     音色名称（默认：zh-CN-XiaoxiaoNeural）
  -r, --rate <1.0>        语速（-50% 到 +50%，默认 0%）
  -p, --pitch <0>         音调（-50Hz 到 +50Hz，默认 0）
  -o, --output <file>     输出音频文件路径
  --list-voices           列出所有可用音色
  --no-send               只生成音频，不发送
  -h, --help              显示帮助

🎵 音色列表

# 查看所有可用音色
bash scripts/send_voice.sh --list-voices

# 查看中文音色
bash scripts/send_voice.sh --list-voices | grep zh-CN

⚙️ 高级配置

1. 自定义默认音色

编辑 config.sh：

DEFAULT_VOICE="zh-CN-YunxiNeural"  # 男声
DEFAULT_RATE="0"                    # 正常语速
DEFAULT_PITCH="0"                   # 正常音调

2. 批量生成

# 从文件读取文字，批量生成
cat messages.txt | while read line; do
    bash scripts/send_voice.sh -t "$line"
    sleep 2
done

📦 文件结构

feishu-edge-tts/
├── SKILL.md
├── README.md
├── reference.md
├── scripts/
│   ├── send_voice.sh        # 主脚本
│   ├── list_voices.sh       # 音色列表
│   └── config.sh            # 配置文件
├── examples/
│   ├── greetings.txt        # 问候语示例
│   └── crontab.txt          # 定时任务示例
└── config.sh                # 用户配置

💰 商业授权

• 个人使用：免费
• 商业使用：请联系作者获取授权

---

Made with ❤️ by 司幼 (SiYou)