Openai Whisper

Local speech-to-text with the Whisper CLI (no API key).

Security Scan

The skill is internally consistent: it declares and uses the whisper CLI, installs via a Homebrew formula that supplies the whisper binary, and its runtime instructions match the stated local transcription purpose.

Purpose & Capability

Name/description, required binary (whisper), and the brew install formula (openai-whisper) all align with a local Whisper CLI wrapper. No unrelated credentials or binaries are requested.

Instruction Scope

SKILL.md only instructs running the whisper CLI against local audio files and notes model cache location (~/.cache/whisper). It does not request unrelated system files, extra environment variables, or external endpoints beyond model downloads on first run.

Install Mechanism

Install uses a Homebrew formula (openai-whisper) to create the whisper binary. Homebrew is a standard package manager and appropriate for installing a CLI. No arbitrary download URLs or archive-extract steps are present in the spec.

Credentials

No environment variables or external credentials are required. The only persisted data noted is model files downloaded to ~/.cache/whisper, which is proportional to a local transcription tool.

Persistence & Privilege

Skill is not forced-always, does not request elevated privileges, and does not modify other skills or system-wide configs. It will run the local whisper binary when invoked (autonomous invocation is platform default).

Assessment

This skill appears to do what it says, but consider these practical checks before installing: (1) Homebrew will install the whisper binary — inspect the openai-whisper formula or install the binary yourself if you prefer to review the source; (2) on first run the CLI will download pretrained models to ~/.cache/whisper — expect large downloads and disk usage and verify network access is acceptable; (3) audio is transcribed locally (no API key required), but the agent invoking the skill can run the CLI on any audio file it has access to — only allow the agent to access files you trust; (4) ensure Homebrew is supported on your OS or install the CLI by hand if necessary; (5) if you want maximal control, preinstall the whisper binary and keep it up-to-date yourself rather than relying on automated installs.

Whisper (CLI)

Use whisper to transcribe audio locally.

Quick start

• whisper /path/audio.mp3 --model medium --output_format txt --output_dir .
• whisper /path/audio.m4a --task translate --output_format srt

Notes

• Models download to ~/.cache/whisper on first run.
• --model defaults to turbo on this install.
• Use smaller models for speed, larger for accuracy.