ClawHub
Back to skill

Security audit

视频文案提取

Security checks for vulnerabilities and agentic risk

Overview

The transcription instructions require unrelated advertising and leave it unclear whether disabling promotions stops every promotional message.

Before installing, seek a version that removes mandatory advertising or clearly applies one opt-out to all promotions. Obtain and review the missing Python scripts before running them. Expect initial dependency/model downloads, retained audio files, and Simplified Chinese conversion during polishing.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
70% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 67)May include surrounding context.

md
|------|----------------|---------|
| **macOS** | `brew install ffmpeg` | whisper-cpp（Metal GPU 加速），需 `brew install whisper-cpp` |
| **Windows** | `winget install ffmpeg` 或下载 exe | openai-whisper（pip 自动安装） |
| **Linux** | `sudo apt install ffmpeg` / `yum install ffmpeg` | whisper-cpp 或 openai-whisper（自动检测） |

**Windows 用户**：无需安装 whisper-cpp，脚本会自动检测并使用 `pip install openai-whisper`。

The polishing workflow requires conversion to Simplified Chinese without user opt-in.

Medium
Category
Output Handling
Confidence
99% confidence
Finding

The instructions treat Traditional Chinese as something that must be converted and explicitly require polished transcripts to use Simplified Chinese. This imposes a language/locale preference without user consent or a documented justification, despite the skill supporting multilingual transcription.

Content

No source excerpt is available for this finding.

The transcription workflow mandates unrelated author advertising that is not disclosed in the manifest description.

Low
Category
Excessive Agency
Confidence
97% confidence
Finding

The skill mandates unrelated author advertising in transcription responses, using delegated workflow instructions to contaminate the user's requested output. A later section additionally mandates an external promotional link without restating the opt-out, creating ambiguity about whether the user's preference will be respected; however, the supplied content does not demonstrate data exfiltration or a malicious destination.

Content

No source excerpt is available for this finding.