WeChat Voice Parsing for Openclaw

A specialized skill for clawbot that decodes WeChat SILK audio files and transcribes them into text using local Whisper models.

jozhn
v0.1.0
Mar 22, 2026
1
938
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install wechat-voice-decode

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install wechat-voice-decode using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is WeChat Voice Parsing?

WeChat Voice is a technical utility designed to bridge the gap between WeChat's proprietary audio formats and actionable AI data. This skill enables an agent to intercept inbound audio attachments, identify the specific SILK_V3 header, and convert the content into standard WAV format.

By leveraging Openclaw Skills, this tool ensures that your AI agent can 'hear' and understand voice messages natively. It uses a local CPU-based Whisper implementation to provide fast, privacy-focused transcriptions without relying on external cloud APIs, making it ideal for responsive and secure WeChat bot integrations.

WeChat Voice Parsing Use Cases

  • Transcribing voice notes from WeChat users into text for AI processing.
  • Analyzing audio attachments to determine if they are silent, too short, or unclear.
  • Providing automated text-based replies based on the content of a voice message.
  • Converting legacy SILK audio files into standard WAV format for archival or debugging.

How WeChat Voice Parsing Works

  1. The agent retrieves the local file path for an inbound audio attachment from the conversation context.
  2. It performs a byte-level inspection to detect the #!SILK_V3 header.
  3. The skill invokes a bundled Python script to decode the SILK stream into a WAV file.
  4. Faster-Whisper processes the WAV file locally on the CPU to generate a text transcript.
  5. The system returns the recognized text or a NO_SEGMENTS flag if the audio is unusable.
  6. The agent generates a user-facing response based on the resulting transcript.

WeChat Voice Parsing Setup

This skill requires Python 3 and FFmpeg to be installed on the host system. To install the necessary dependencies for these Openclaw Skills, run:

python3 -m pip install --user silk-python faster-whisper

Ensure that the bundled transcription script is accessible within your workspace path.

WeChat Voice Parsing Data Schema & Taxonomy

The skill manages audio data and metadata through the following structure:

Data Type Description
Input File Raw WeChat SILK audio (detected via hex header #!SILK_V3)
Intermediate Temporary WAV file generated during the decoding phase
Output Plaintext transcript or standardized status code (NO_SEGMENTS)
Scripts transcribe_wechat_voice.py serves as the primary execution entry point

WeChat Voice Parsing Advanced Features

  • Local CPU Optimization: Uses faster-whisper to avoid heavy GPU/CUDA dependencies while maintaining high accuracy.
  • Format Validation: Includes byte-level header inspection to prevent processing errors on non-SILK files.
  • Error Resilience: Explicit handling for silent or extremely short clips to provide better user feedback.
  • Integrated Debugging: Supports manual hex inspection and byte-size verification for troubleshooting attachment issues.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*