Local Vosk STT
PassAudited by ClawScan on May 1, 2026.
Overview
The skill appears purpose-aligned for local speech-to-text, with a review note because setup depends on external downloads and a referenced helper script not included in the artifact.
This looks like a benign local transcription helper, but install it carefully: confirm the missing transcribe helper is from a trusted source, verify the Vosk package/model downloads, and be aware that metadata does not declare all local tools implied by the instructions.
Findings (1)
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.
The setup may install local Python code and model files from external sources, and the documented helper script is not reviewable in the provided artifact.
The skill relies on a local helper path plus external package/model downloads. These are plausible for Vosk transcription, but the helper is not present in the provided manifest and the downloads are not pinned by hash.
./skills/local-vosk/scripts/transcribe audio.mp3 ... pip3 install vosk --user --break-system-packages ... wget https://alphacephei.com/vosk/models/vosk-model-small-en-us-0.15.zip
Before using it, verify that any local transcribe script exists and comes from the expected source, and prefer official or pinned downloads for the Vosk package and model.