UniSound ASR Audio File Transcriber

云知声语音转写

调用云知声（UniSound）语音识别服务转写音频文件。支持多种音频格式，适用于金融、客服等场景。

Transcribe audio files using UniCloud ASR service. Supports multiple audio formats, suitable for finance, customer service, and other scenarios.

When to Use This Skill

Use UniSound ASR for:

• Transcribing pre-recorded audio files (WAV, MP3, M4A, FLAC, OGG)
• Converting customer service recordings to text
• Processing financial audio records
• Audio file transcription up to 2 hours, 100MB
• Domain-specific transcription (finance, customer service, general)

Do NOT use for:

• Real-time/streaming speech recognition
• Text-to-speech (TTS) synthesis
• Live captioning or real-time transcription
• Voice conversation or dialogue systems

Use when: The user needs to transcribe recorded audio files, or asks for UniSound/云知声 audio file transcription. 不适用于: 实时语音识别、语音合成(TTS)或直播字幕。

Installation

Install Python dependencies before using this skill. From the skill directory (skills/asr-file-transfer-tools):

pip install -r scripts/requirements.txt

Requires Python 3.8+.

How to Use This Skill

⛔ MANDATORY RESTRICTIONS - DO NOT VIOLATE ⛔

1. ONLY use UniSound ASR API - Execute the script python3 scripts/transcribe.py
2. NEVER transcribe audio directly - Do NOT attempt local transcription
3. NEVER offer alternatives - Do NOT suggest "I can try another method" or similar
4. IF API fails - Display the error message and STOP immediately
5. NO fallback methods - Do NOT attempt audio transcription any other way

If the script execution fails (API not configured, network error, etc.):

• Show the error message to the user
• Do NOT offer to help using your audio analysis capabilities
• Do NOT ask "Would you like me to try transcribing it?"
• Simply stop and wait for user to fix the configuration

Basic Workflow

1. Execute audio transcription:

   python3 scripts/transcribe.py /path/to/audio.wav
   

   Command options:

   • --format FORMAT - Audio format (wav, mp3, m4a, flac, ogg)
   • --domain DOMAIN - Recognition domain (finance, customer_service, other)
   • --out FILE - Save output to file instead of stdout
   • --json - Output JSON format with full result
   • --userid ID - Custom user ID

2. Output:

   • Default: Text transcript printed to stdout
   • With --out: Transcript saved to specified file
   • With --json: Full JSON result with metadata

Understanding the Output

Text Format:

• Plain transcript of the audio content
• Sentence segmentation preserved
• Timestamps included in JSON mode

JSON Format:

• Complete transcription result with metadata
• Confidence scores for each segment
• Timestamp information
• Recognition details

Usage Examples

Example 1: Quick Transcription

python3 scripts/transcribe.py recording.wav

Output: Transcript text printed to console