Security audit

Browser Automation Zh

Security checks for vulnerabilities and agentic risk

Overview

This browser automation skill is coherent, but it encourages broad scraping, automated login, screenshots, and form submission without clear safety boundaries.

Review this skill before installing if you will use it on third-party sites, logged-in dashboards, or production systems. Use it only where you have authorization, avoid storing secrets or sensitive screenshots in plaintext, confirm before submitting forms, and keep any scheduled automation explicit and easy to disable.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The skill description is written broadly enough to attract generic requests for scraping, browser automation, and monitoring, which increases the chance it will be invoked for sensitive or policy-relevant actions without sufficient narrowing. In this context, broad triggering is risky because the skill enables scalable interaction with third-party sites and data collection, and the documentation does not define boundaries such as authorization, rate limits, or acceptable targets.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The documentation explicitly promotes automated login, data capture, form submission, and screenshot saving, but provides no safety guidance for handling credentials, personal data, session tokens, or operational impact on target systems. This is dangerous because browser automation can easily access privileged dashboards, submit unintended actions, or capture sensitive information in screenshots and logs, especially when used against production systems or third-party services.

Natural-Language Policy Violations

Low

Confidence: 78% confidence
Finding: The skill name includes a Chinese locale marker and all descriptive and usage content is presented in Chinese, with no indication that other languages are supported or that Chinese-only behavior is intentional and justified. This may violate language or locale choice expectations if the organization requires user opt-in for language constraints.

Static analysis

No suspicious patterns detected.