openclaw-whisper-asr

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward local Whisper speech-to-text setup, with a privacy hygiene caveat for voice files and transcripts.

Install only on a machine where you are comfortable installing packages and building whisper.cpp. Pin the upstream repository to a trusted release or commit where possible, use unique restricted temporary filenames instead of a shared /tmp/audio.wav path, and delete audio/transcript files after processing if they may contain private information.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill explicitly targets converting voice messages from platforms like Feishu and Telegram into text, but it provides no warning about the sensitive nature of user audio or the privacy implications of storing, processing, or exposing transcripts. Voice messages often contain personal, confidential, or regulated information, so omitting handling guidance increases the risk of accidental privacy violations, over-retention, and unauthorized access to transcripts or temporary audio files.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal