Wjs Transcribing Audio

• Outcome: Review
• Latest audit: May 28, 2026, 2:30 AM
• Version:
  0.1.0

Overview

The skill appears to be a transcription tool, but it can send user audio to external speech services using local API keys without clearly requiring user approval first.

Install only if you are comfortable with audio files being processed by external transcription providers and with the skill using local API keys. Prefer configuring only the required provider key in a dedicated environment file, and ask the agent to confirm before sending any confidential, private, or regulated audio.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The skill explicitly instructs loading credentials from `~/code/.env` and using them for third-party ASR services, but it provides no user-facing notice that local secrets and user audio will be sent to external providers. In an agent setting, this can cause silent use of sensitive API keys and external processing of potentially confidential media without informed consent or clear scope controls.

External Transmission

Medium

• Category: Data Exfiltration

Content

"temperature": "0.2",                    # enable fallback chain (anti-loop)
}
with open("chunk.mp3", "rb") as f:
    r = httpx.post(
        "https://api.openai.com/v1/audio/transcriptions",
        headers={"Authorization": f"Bearer {os.environ['OPENAI_API_KEY']}"},
        data=data,

• Confidence: 93% confidence
• Finding: httpx.post( "https://

VirusTotal

VirusTotal findings are pending for this skill version.