ClawHub
Back to skill

Security audit

Whisper语音转文字专业版

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a coherent Whisper transcription guide, but its broad activation language and powerful read/write/command/API capabilities need user review before installation.

Install only if you want an agent to handle local audio/video transcription workflows and related setup commands. Before use, narrow activation to explicit transcription/subtitle/diarization tasks, review any dependency installs, avoid processing confidential recordings through an exposed API without authentication and retention controls, and confirm where transcripts, logs, caches, and temporary files are written.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
83% confidence
Finding
The skill declares an LLM API as a required dependency even though the described Whisper transcription workflow is primarily local and non-LLM. Unnecessary external AI dependency can cause sensitive audio, transcripts, or metadata to be routed to third-party services without clear justification, expanding the data exposure and trust boundary.

Vague Triggers

High
Confidence
89% confidence
Finding
The trigger conditions are so broad that the skill may activate for generic productivity or automation requests outside the user's actual intent. In a skill with read/exec/write capabilities, overbroad activation increases the chance of inappropriate file processing, command execution, or handling of sensitive audio data without sufficiently specific user consent.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The API example accepts uploaded audio, writes it to a temporary file, processes it, and returns transcript data, but it lacks clear user-facing warnings about sensitive content handling, retention, and exposure risks. In enterprise transcription scenarios, audio and transcripts commonly contain confidential or regulated information, so omission of those safeguards materially increases privacy and compliance risk.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.