persona-voice

ReviewAudited by ClawScan on May 1, 2026.

Overview

The skill does what it claims—turning Feishu/Lark bot replies into SenseAudio voice messages—but it requires bot/API credentials and sends voice/text data to external services.

Install only if you are comfortable giving it Feishu bot credentials and a SenseAudio API key, and if sending conversation audio/text through those services fits your privacy requirements. Restrict the bot to intended chats, protect the secrets, and use a trusted ffmpeg installation.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Medium

#ASI02: Tool Misuse and Exploitation

What this means

The bot can send voice messages into Feishu chats, so mistakes or unwanted invocations could post content to real conversations.

Why it was flagged

The skill can post an audio message into a Feishu chat using a supplied chat_id. This is the core feature and is disclosed, but it is still an external message-sending action.

Skill content

requests.post(f'{self.base_url}/open-apis/im/v1/messages', params={'receive_id_type': 'chat_id'}, ... json={'receive_id': chat_id, 'msg_type': 'audio', ...})

Recommendation

Use a least-privilege Feishu bot, restrict where it is installed, and make sure the chat_id and generated reply are appropriate for the target chat.

Medium

#ASI03: Identity and Privilege Abuse

What this means

Anyone running the skill must provide credentials that can authenticate API calls and send messages through the configured bot.

Why it was flagged

The skill requires Feishu bot credentials and a SenseAudio API key. This is purpose-aligned, but it grants delegated access to third-party services.

Skill content

export FEISHU_APP_ID="cli_xxx"
export FEISHU_APP_SECRET="your_secret"
export SENSEAUDIO_API_KEY="your_key"

Recommendation

Store these secrets securely, rotate them if exposed, and grant only the permissions required for audio upload and message sending.

Medium

#ASI07: Insecure Inter-Agent Communication

What this means

User audio, transcribed text, and generated replies may leave the local environment and be processed by external services.

Why it was flagged

The stated workflow sends audio/text to SenseAudio for transcription/synthesis and sends the resulting voice message through Feishu. These provider flows are disclosed and central to the skill.

Skill content

ASR / TTS：统一使用 SenseAudio。 ... 飞书回复：统一发送为飞书原生语音条

Recommendation

Avoid using the skill for highly confidential conversations unless SenseAudio and Feishu handling of that data is acceptable.

Low

#ASI05: Unexpected Code Execution

What this means

A local ffmpeg executable must be available and will process the generated audio file.

Why it was flagged

The skill runs the local ffmpeg binary to convert generated audio to OPUS. This is expected for Feishu native voice messages and is not executed through a shell.

Skill content

cmd = [ffmpeg_bin, '-y', '-i', str(src), '-c:a', 'libopus', '-b:a', '32k', ...]
proc = subprocess.run(cmd, capture_output=True, text=True, check=False, timeout=120)

Recommendation

Install ffmpeg from a trusted source and set FFMPEG_PATH only to a trusted binary.