Feishu Voice Chat
v0.1.1飞书语音对话能力，提供语音识别（ASR）和语音合成（TTS）功能， 所有的飞书语音消息都通过该技能处理。 完整语音交互链路：接收用户语音 → ASR 转文字 → LLM 处理 → TTS 转语音 → 通过飞书插件发送语音消息。 当用户要求"语音回复/说给我听"时，只回复飞书语音消息（audio 气泡），不回复文本...
⭐ 0· 108·0 current·0 all-time
byneo@li-neo
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description, declared env vars (VOLC_*), SKILL.md usage, and the two scripts all align: they implement ASR and TTS via Volcengine and return an agent message command to send audio via the Feishu plugin. One minor inconsistency: the scripts call ffmpeg for format conversion but the skill metadata lists no required binaries; ffmpeg should be declared as a required binary.
Instruction Scope
SKILL.md instructs running the included scripts with local audio paths. The scripts only read audio files from /tmp/openclaw media paths (and other allowed temp dirs), load a local .env for VOLC credentials, call Volcengine endpoints for ASR/TTS, optionally convert with ffmpeg, and emit a structured message_cmd for the agent to call the message tool. They do not attempt to read unrelated user files or exfiltrate arbitrary data beyond sending audio/text to the declared TTS/ASR endpoints. The scripts also implement path checks to limit accessible paths.
Install Mechanism
This is an instruction-only skill with included Python scripts and a small requirements.txt (requests, python-dotenv). There is no remote download or archive extraction. Installation risk is low. Note: the skill relies on an external binary (ffmpeg) but does not declare it in the metadata.
Credentials
The only required environment variables are VOLC_APPID and VOLC_ACCESS_TOKEN (plus optional VOLC_* settings) which are appropriate for the declared Volcengine ASR/TTS integration. No unrelated credentials or broad secrets are requested. The scripts load a local .env from the skill directory and also permit reading these env vars from the environment, which is expected for service auth.
Persistence & Privilege
The skill does not request always: true and does not modify other skills or global agent configuration. It writes temporary audio files under /tmp/openclaw and relies on normal subprocess calls; privileges are limited to the agent's runtime environment.
Assessment
This skill appears to do exactly what it claims: convert incoming Feishu audio to text with Volcengine ASR and synthesize audio with Volcengine TTS, then return an agent 'message' command to send the audio. Before installing: 1) Be aware that all audio and text are sent to Volcengine endpoints (Bytedance domains shown in the scripts) — ensure you are comfortable with that privacy/third‑party processing. 2) Provide only the Volcengine credentials (VOLC_APPID and VOLC_ACCESS_TOKEN) required; do not reuse broader platform or AWS/GCP credentials. 3) Install ffmpeg on the host (the scripts call ffmpeg for format conversion but the skill metadata did not list it as a required binary). 4) Confirm the skill is installed where the scripts expect (.openclaw/workspace/skills/feishu-voice-chat or adjust SCRIPT_PATH) or edit the wrapper script accordingly. 5) Review .env placement and file permissions so the VOLC tokens stay protected. If you need higher assurance, audit network traffic to the declared endpoints and inspect logs to confirm only audio/text related to user messages are transmitted.Like a lobster shell, security has layers — review code before you run it.
latestvk974hzs1f3dqnmerkb312cwbcd83dqk7
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Environment variables
VOLC_APPIDrequired— 火山引擎应用 IDVOLC_ACCESS_TOKENrequired— 火山引擎访问令牌VOLC_RESOURCE_IDoptional— TTS 资源 ID (默认: seed-tts-1.0)VOLC_VOICE_TYPEoptional— TTS 音色代码VOLC_RESOURCE_ID_ASRoptional— ASR 资源 ID