ClawHub
Back to skill

Security audit

Browser Use

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed browser-automation guide with no executable install payload, but users should apply it carefully for scraping and logged-in sites.

Install this only if you want an agent to control a browser for browsing, clicking, scrolling, and extracting page data. Avoid using it to collect sensitive, private, paywalled, or restricted data, and review site rules before bulk scraping.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The activation text is broad enough to trigger on many ordinary browsing or information-gathering requests, increasing the chance the agent invokes powerful browser automation when a narrower skill or direct answer would suffice. Over-broad activation expands the operational surface for web interaction, scraping, and unintended site actions, which can lead to mis-execution or policy-bypassing behavior in ambiguous contexts.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill explicitly supports web scraping, data collection, and batch extraction but does not include guardrails around privacy-sensitive data, terms-of-service constraints, rate limiting, or impact on target sites. In context, this omission makes the browser automation capability more dangerous because it can be used at scale without prompting the agent to assess whether collection is appropriate or whether personal or restricted data may be involved.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.