Chrome Automation

Security checks across malware telemetry and agentic risk

Overview

This skill provides useful browser automation, but it also enables stealth anti-bot browsing and automated login/form submission without enough safeguards.

Install only if you intentionally need headless browser automation and can run it in an isolated, low-privilege environment. Do not use it to bypass third-party anti-bot controls without authorization, and avoid using real credentials until logging is removed and explicit approval is required before login, posting, or form submission.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The code deliberately injects stealth JavaScript to hide automation indicators such as navigator.webdriver, spoof plugins, and fabricate window.chrome. In the context of a skill explicitly advertising bypass of anti-bot detection, this is not ordinary browser automation but intentional evasion of site security controls, enabling deceptive scraping, account abuse, or automated actions against protected services.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill explicitly supports automated form filling and login but does not warn users that credentials, cookies, or other sensitive inputs may be handled by browser automation and potentially exposed to target sites or logs. In this context, the omission is security-relevant because users may run login automation without understanding the privacy and account-risk implications.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The documentation encourages browser automation against external websites but does not disclose that running these commands will make outbound network requests and may transmit user-entered data, cookies, browser fingerprints, and other metadata. Because the skill also advertises stealth and anti-detection features, the lack of notice increases the chance of unintended data disclosure or misuse.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The form automation function can navigate to an arbitrary URL, fill arbitrary selectors, and optionally click submit with no confirmation, allowlist, or dry-run safeguard. That enables user-affecting operations such as account changes, purchases, postings, or consent flows to be triggered automatically if this helper is invoked with untrusted or mistaken inputs.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal