Voice Wake Say
PassAudited by ClawScan on May 1, 2026.
Overview
This instruction-only skill appears benign: it narrowly tells the assistant to speak voice-triggered responses aloud with macOS `say` and does not request credentials, networking, code installation, or persistence.
This skill is suitable if you want Voice Wake messages read aloud locally on macOS. Be aware that responses may be audible to people nearby, and the registry metadata does not declare the macOS/`say` dependency even though the instructions rely on it.
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.
When the exact Voice Wake trigger is present, the assistant may speak its response aloud through the Mac's speakers.
The skill directs the assistant to invoke the local macOS `say` command. This is disclosed and central to the skill's purpose, but users should notice that installing it authorizes local command use for text-to-speech.
printf '%s' "$SPOKEN_TEXT" | say
Use it only if you want local speech output, and avoid voice-triggered use for sensitive responses when others may overhear.
On non-macOS systems, or systems without `say`, the spoken-output feature may fail even though the registry does not pre-warn about that requirement.
The skill documentation is macOS-specific and uses `say`, but the registry metadata does not declare the OS or binary dependency. This is a compatibility/declaration gap rather than evidence of hidden behavior.
OS restriction: none; Required binaries (all must exist): none
Treat this as a macOS-only skill and confirm that `say` is available if you expect audio output.