Security audit

cloakbrowser-stealth

Security checks for vulnerabilities and agentic risk

Overview

This skill is transparently a stealth browser scraping tool, but its anti-detection, proxy guidance, and unconstrained local screenshot writing are high-impact enough to require careful review before use.

Install only if you are authorized to automate the target sites and you accept the risk of using stealth and proxy-based scraping. Avoid logged-in or personal pages, treat screenshots and browser profiles as sensitive local data, choose explicit temporary output paths, clean up captures after use, and review the CloakBrowser package before relying on it.

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Description-Behavior Mismatch

Medium

Confidence: 87% confidence
Finding: The script accepts a user-controlled --screenshot path and writes page content to that location without validation or confinement. In an agent/tooling context, this enables arbitrary local file writes within the agent's permissions, which expands the tool from read/fetch behavior into persistent disk modification and can overwrite sensitive files or deposit scraped data in unsafe locations.

Missing User Warnings

Medium

Confidence: 85% confidence
Finding: The skill explicitly encourages saving screenshots of protected pages to disk but provides no warning that screenshots may capture sensitive or copyrighted content and leave it in predictable local storage. In a stealth-scraping context, this increases the chance of retaining sensitive page data without user awareness or cleanup, creating confidentiality and data-handling risk.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The file documents stealth anti-detection features, persistent profiles, and proxy-based geolocation matching in a way that facilitates evasion of website bot defenses while omitting any safety, privacy, or abuse constraints. In the context of a skill explicitly intended to bypass anti-bot protections, the lack of warnings about account bans, data leakage via persistent profiles, or misuse of proxies increases the likelihood of unsafe or unauthorized use.

Natural-Language Policy Violations

Medium

Confidence: 80% confidence
Finding: The language describes spoofing navigator language settings to appear normal, which is a deception-oriented anti-detection capability. Within a skill whose stated purpose is to bypass bot detection on protected sites, undocumented locale spoofing can be used to evade fraud controls and can misrepresent client identity without user awareness or justification.

Static analysis

No suspicious patterns detected.