ClawHub
Back to skill

Security audit

短视频生成专业版

Security checks across malware telemetry and agentic risk

Overview

The skill is mainly a short-video generation workflow, but it gives broad read/write/command authority with activation wording and capability claims that are not tightly limited to video tasks.

Install only if you intend to let the agent read and write project video files and run local video-generation commands. Keep use limited to explicit video-production tasks, review any generated shell commands before execution, configure TTS or platform credentials through environment variables or a secret manager, and avoid running privileged setup commands outside isolated CI or a prepared environment.

SkillSpector

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The skill advertises broad capabilities like generic API integration, command execution, file handling, and information retrieval that go beyond the narrowly described video-generation purpose. Overbroad capability claims can cause the agent to invoke this skill in unrelated contexts, increasing the chance of unnecessary file access, external communication, or command execution under a misleading business-use cover.

Intent-Code Divergence

Medium
Confidence
87% confidence
Finding
The documentation says no additional API key is needed while also stating that LLM and TTS services are required and may need cloud configuration. This inconsistency can mislead operators into enabling external services without proper credential handling review, increasing the risk of accidental secret exposure or unsafe assumptions about local-only execution.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The activation wording is very broad, including general productivity and automation themes rather than a tightly bounded video-generation trigger. That makes accidental or excessive activation more likely, which is dangerous here because the skill has read, write, and exec tools and may run commands or manipulate files in situations unrelated to its intended purpose.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The capability coverage section includes vague trigger keywords such as 'Use' and 'when', which are not meaningful security boundaries and can match many unrelated prompts. In a skill with execution and file-writing abilities, ambiguous trigger scope materially raises the risk of unintended invocation and overbroad agent behavior.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Content
steps:
      - uses: actions/checkout@v3
      - name: Setup FFmpeg
        run: sudo apt install ffmpeg
      - name: Batch Generate Videos
        run: |
            --tasks tasks.json \
Confidence
78% confidence
Finding
The CI/CD example includes 'sudo apt install ffmpeg', which normalizes privileged package installation inside an execution-capable skill. While common in GitHub Actions examples, it encourages root-level command execution and could be dangerous if copied into less isolated environments or combined with modified package sources or additional injected commands.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.