Meeting Transcriber

基于FunASR的中文实时语音转录，支持麦克风录音、自动标点、智能VAD及转录文件管理与环境检测。

This skill appears to do what it says, but it is tailored to a specific Windows/Conda setup and uses hard-coded paths (e.g., D:\dev\python\voiceFunAsr, D:\ProgramData\Miniconda3\envs\audioProject, C:\Users\pengjschina\.cache). Before installing or running: 1) inspect vocie_mic_fixed_gbk.py and any referenced scripts in the D: path — those are the actual transcription engine and could contain arbitrary code; 2) be aware the skill will create startup scripts and a config file under your home (~/.openclaw) and may modify SKILL.md; 3) it will spawn subprocesses (conda, python) and may download ASR models at first run — ensure network/download behavior is acceptable; 4) if you do not have a matching Windows/Conda environment, the skill will likely fail or attempt to modify files to adapt; 5) no secrets are requested by the skill, but only proceed if you trust the source of the included transcriber scripts.

Type: OpenClaw Skill Name: meeting-transcriber Version: 1.0.0 The skill bundle provides a legitimate meeting transcription service using the FunASR engine. It includes comprehensive scripts for environment validation (check_env.py), Conda configuration (conda_setup.py), and process management (transcriber.py). While the code utilizes high-privilege operations such as subprocess execution to launch shells and psutil to terminate processes, these actions are strictly aligned with the functional requirements of managing real-time audio transcription on Windows. The presence of hardcoded paths and specific user directories (e.g., 'pengjschina') indicates the tool was developed for a specific local environment, but there is no evidence of malicious intent, data exfiltration, or unauthorized remote access.