ClawHub

Simple sound-to-text skill locally

v1.0.2

Local speech-to-text using OpenAI Whisper. Use when the user needs to: (1) transcribe audio files to text, (2) convert voice messages to written content, (3)...

⭐ 0· 179·0 current·0 all-time
byKuikui@lkisme
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description (local Whisper STT) match the included scripts and instructions. The install script, Python transcription script, model choices, and output locations are all consistent with the stated purpose.
✓
Instruction Scope
SKILL.md and the scripts only reference local audio files, creating a local venv, installing dependencies, and writing outputs under /root/.openclaw/workspace/stt_output (and models to ~/.cache/whisper). There are no instructions to read unrelated system files, pull secrets, or send transcription data to external endpoints.
ℹ
Install Mechanism
Installation is via a provided shell script that creates a Python venv, runs pip install openai-whisper, and may call system package managers (apt-get/yum/brew) to install ffmpeg. This is a common approach but requires network access to package registries and privileges to install system packages. The install script itself is included (no hidden download URLs).
✓
Credentials
The skill requests no environment variables or credentials. Its filesystem paths are under /root/.openclaw and standard caches; this matches local STT behavior and does not require unrelated secrets.
✓
Persistence & Privilege
Skill is not always-enabled and does not modify other skills or global agent configuration. It creates its own venv and output directories under /root/.openclaw, which is normal for an installed skill.
Assessment
This skill appears to do what it claims: local Whisper-based transcription. Before installing, be aware that the included install.sh will: (1) create a virtual environment under /root/.openclaw/venv/stt-simple, (2) run pip install openai-whisper (downloads packages from PyPI), (3) attempt to install FFmpeg via apt/yum/brew (requires sudo/root), and (4) trigger model downloads to ~/.cache/whisper (large files, network bandwidth). If you plan to run this on a shared system, consider running the install and transcription inside a container or VM, and review the script contents (they are included) before execution. No credentials are requested and there are no hidden external endpoints in the code, but only run third-party install scripts that you trust or in an isolated environment.

Like a lobster shell, security has layers — review code before you run it.

latestvk974p129qyezn5qzrcr59vyytn82qkts

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments