ClawHub
Back to skill

Security audit

SurfAgent

Security checks for vulnerabilities and agentic risk

Overview

This browser-automation skill appears purpose-aligned, but it gives an agent high-impact access to real logged-in browser sessions without enough safety scoping.

Install only if you intentionally want an agent to operate a real logged-in browser profile. Prefer a dedicated isolated browser profile, avoid personal or production accounts unless explicitly needed, and require approval before form submission, purchases, account changes, cookie access or changes, JavaScript execution, and bulk data extraction.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The README advertises full control of a real, persistent Chrome profile, including cookies, arbitrary JavaScript execution, crawling, extraction, and bot-detection bypass, but provides no safety warning about access to authenticated sessions, sensitive data, or the risk of destructive actions in the user's browser state. In this context, omission of those warnings is dangerous because users may install and authorize the skill without understanding that an agent could read private content, act on logged-in accounts, modify browser state, or exfiltrate session data.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill explicitly advertises use of a persistent Chrome profile with real cookies and sessions, which gives the agent access to authenticated browser state and potentially sensitive personal or corporate accounts. Without a prominent warning, scoped-consent guidance, or restrictions on which sites/actions are allowed, users may unknowingly expose sessions, tokens, browsing history, and private data to automated actions.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The listed capabilities include clicking, typing, form submission, cookie manipulation, and arbitrary JavaScript execution in a real browser session, all of which can change account state, submit transactions, alter settings, or exfiltrate data from authenticated pages. In the context of a persistent browser profile, these are not merely read-only automation features and should be treated as high-risk actions requiring explicit safeguards.

Static analysis

No suspicious patterns detected.