ClawHub
Back to skill

Security audit

豆包批量视频生成

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent Doubao video-generation workflow with disclosed browser-session and local video-processing risks.

Install only if you are comfortable letting the agent operate an authenticated Doubao browser session, consume generation quota, download generated videos, and write media files locally. Use a dedicated browser profile for Doubao, review prompts before generation, and avoid uploading sensitive local files.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill instructs use of shell commands and file-writing behavior (ffmpeg, python, concat-list.txt creation, output files) but does not declare corresponding permissions. Undeclared capabilities undermine informed consent and policy enforcement, increasing the chance that an agent can write files or execute commands unexpectedly in the user's environment.

Tp4

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding
The documented behavior claims end-to-end browser automation, prompt generation, login handling, and batch video creation, but the concrete executable parts shown primarily cover post-processing of local files. This mismatch is dangerous because users and reviewers may trust the skill with browser sessions, credentials, and automation assumptions that are not actually implemented or auditable, masking real behavior and weakening security review.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The trigger phrases include broad natural-language requests such as '帮我用豆包做视频' and '批量生成短视频', which may activate the skill in situations the user did not intend. Unintended activation is especially risky here because the skill can lead to browser automation, persistent login use, shell execution, and file generation.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill describes maintaining login state via a persistent browser profile and user data directory, but does not prominently warn users that browser automation will retain authentication artifacts and other browsing data. Persistent session handling increases exposure of account access, cookies, and local profile data if the environment is shared, compromised, or reused unexpectedly.

Static analysis

No suspicious patterns detected.