MH openai-whisper
PassAudited by ClawScan on May 1, 2026.
Overview
This skill coherently wraps the local Whisper speech-to-text CLI and discloses its required install and model cache behavior.
This appears safe for its stated purpose: local speech-to-text. Before installing, confirm you are comfortable installing the Homebrew Whisper package and allowing Whisper to download model files into ~/.cache/whisper on first use.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Installing the skill may require installing the Whisper CLI through Homebrew.
The skill relies on an external Homebrew package to provide the Whisper CLI. This is disclosed and central to the skill's purpose, but users should recognize it installs third-party software.
"install": [{ "id": "brew", "kind": "brew", "formula": "openai-whisper", "bins": ["whisper"] }]Install only if you are comfortable with the Homebrew openai-whisper package and keep it updated through normal package-management practices.
The first transcription may download model files and use disk space in the local Whisper cache.
The skill discloses that Whisper downloads model files and stores them locally. This is expected for local Whisper transcription, but it is a persistent external asset download.
Models download to `~/.cache/whisper` on first run.
Be aware of the first-run download and local cache location, especially on limited storage or restricted-network systems.