Whisper语音转文字免费版

• Outcome: Review
• Latest audit: Aug 29, 2026 · 6:43 AM UTC
• Version: 1.0.1

Overview

This is mostly a local Whisper transcription guide, but it has inconsistent API-key guidance and an overbroad trigger that could cause exec-enabled agents to use it for unrelated media tasks.

Install only if you want a local command-line transcription helper and are comfortable reviewing the setup commands yourself. Do not provide API keys for this skill based on the current text, and keep usage limited to explicit transcription, subtitle generation, or speech translation tasks rather than general media editing or dubbing.

SkillSpector

Vulnerability Patterns

• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Tool Misuse: Tool Parameter Abuse, Chaining Abuse, Unsafe Defaults
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (7)

Intent-Code Divergence

Medium

• Confidence: 95% confidence
• Finding: The documentation contradicts itself by first claiming the skill is fully local and requires no API key, then later stating that an API key is required. This can mislead an agent or user into unnecessary credential handling, increasing the chance of secret collection, storage, or disclosure contrary to the tool’s actual design.

Intent-Code Divergence

Medium

• Confidence: 93% confidence
• Finding: The security notes discuss API-key leakage and HTTPS/SSL protections even though the skill is described as a local offline transcription tool. These inconsistent security claims can cause agents to assume nonexistent network/API behavior, potentially prompting unnecessary secret usage or unsafe network troubleshooting steps.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The trigger condition is overly broad, covering generic video processing, audio editing, media conversion, and dubbing requests that exceed the stated Whisper transcription scope. In an agent context, this can cause inappropriate auto-invocation of exec-capable behavior for unrelated tasks, increasing the risk of unintended command execution or file manipulation.

Missing User Warnings

Low

• Confidence: 84% confidence
• Finding: The skill describes output generation and automatic model downloads but does not prominently warn that it will create local files and consume disk space. In an automated environment, this can lead to unexpected writes, storage exhaustion, or user confusion about where artifacts and cached models are stored.

Sudo/Root Execution

Medium

• Category: Privilege Escalation

Content

brew install ffmpeg

# Ubuntu / Debian
sudo apt update && sudo apt install ffmpeg

# Windows(推荐使用 scoop 或 choco)
scoop install ffmpeg

• Confidence: 91% confidence
• Finding: The duplicated finding points to the same privileged installation command using sudo. The risk remains that an exec-enabled agent or inattentive user could perform system-level package installation with elevated privileges based solely on skill guidance.

Sudo/Root Execution

Medium

• Category: Privilege Escalation

Content

brew install ffmpeg

# Ubuntu / Debian
sudo apt update && sudo apt install ffmpeg

# Windows(推荐使用 scoop 或 choco)
scoop install ffmpeg

• Confidence: 91% confidence
• Finding: The duplicated finding points to the same privileged installation command using sudo. The risk remains that an exec-enabled agent or inattentive user could perform system-level package installation with elevated privileges based solely on skill guidance.

Chaining Abuse

High

• Category: Tool Misuse

Content

brew install ffmpeg

# Ubuntu / Debian
sudo apt update && sudo apt install ffmpeg

# Windows(推荐使用 scoop 或 choco)
scoop install ffmpeg

• Confidence: 90% confidence
• Finding: The chained command `sudo apt update && sudo apt install ffmpeg` combines multiple privileged operations into a single step, reducing opportunities for review and making accidental or automated execution more dangerous. In an exec-capable skill, command chaining amplifies the blast radius because one approval can trigger several system-modifying actions.

VirusTotal

63/63 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.