← Back to Skills Marketplace
13956
Downloads
12
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install local-whisper
Description
Local speech-to-text using OpenAI Whisper. Runs fully offline after model download. High quality transcription with multiple model sizes.
Usage Guidance
Install it as an unprivileged user, expect a large initial model/dependency download, and consider pinning or reviewing package versions if you need reproducible or tightly controlled installs.
Capability Analysis
Type: OpenClaw Skill
Name: local-whisper
Version: 1.0.0
The skill bundle provides local speech-to-text functionality using OpenAI Whisper. The `SKILL.md` clearly outlines its purpose, usage, and setup, including legitimate dependencies like `ffmpeg` and Python libraries `openai-whisper` and `torch` from trusted sources (download.pytorch.org). The `scripts/transcribe.py` script correctly implements the transcription logic without any signs of data exfiltration, malicious execution, persistence, or prompt injection against the agent. The only network activity is the expected initial download of Whisper models by the `openai-whisper` library, which is consistent with the 'offline after model download' description.
Capability Assessment
Credentials
The skill needs ffmpeg, local model files, and access to the chosen audio file, which is proportionate for offline transcription after the initial model download.
Install Mechanism
Setup uses a uv-managed .venv and installs click, openai-whisper, and torch from a documented HTTPS package index, but the dependencies are not pinned or hash-locked.
Instruction Scope
Runtime behavior is narrow and user-directed, but the usage examples reference scripts/local-whisper while the inspected artifact contains scripts/transcribe.py; this looks like a packaging/usability mismatch rather than hidden security behavior.
Persistence & Privilege
No services, startup hooks, credential access, memory writes, privilege escalation, destructive actions, or cross-session backdoors were found; persistence is limited to the skill files, virtual environment, and normal model cache behavior.
Purpose & Capability
The stated purpose is local Whisper speech-to-text, and the bundled script loads a selected Whisper model, transcribes a user-provided audio file, and prints text or JSON output.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install local-whisper - After installation, invoke the skill by name or use
/local-whisper - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of local-whisper: local speech-to-text using OpenAI Whisper, fully offline after model download.
- Supports multiple model sizes for different speed/quality needs: tiny, base (default), small, turbo, large-v3.
- Includes options for language selection, timestamps, JSON output, and quiet mode.
- Provides clear setup instructions using uv-managed Python virtual environment.
- Requires ffmpeg for audio processing.
Metadata
Frequently Asked Questions
What is Local Whisper?
Local speech-to-text using OpenAI Whisper. Runs fully offline after model download. High quality transcription with multiple model sizes. It is an AI Agent Skill for Claude Code / OpenClaw, with 13956 downloads so far.
How do I install Local Whisper?
Run "/install local-whisper" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Local Whisper free?
Yes, Local Whisper is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Local Whisper support?
Local Whisper is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Local Whisper?
It is built and maintained by araa47 (@araa47); the current version is v1.0.0.
More Skills