ClawHub
Back to skill

Security audit

Browser Read

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward browser page text extractor, with the main caution that it can read content from authenticated pages the user opens.

Install only if you are comfortable letting the agent read the visible text and URL of pages you open for it. Avoid using it on banking, email, admin, HR, private internal, or other highly sensitive pages unless sharing that page content with the agent is intentional.

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
  • 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 (2)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This skill explicitly targets pages that may only be accessible through the user's authenticated browser session, including auth-required pages and social platforms. Without a clear warning about the possibility of extracting sensitive in-session content, users may invoke it on pages containing private account data, internal resources, or other protected information and unintentionally expose that content to the agent workflow.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This script extracts and returns the full visible page content, title, author metadata, URL, and site name from whatever page is currently open in the browser. In the stated skill context, that can include authenticated or bot-protected pages, so it may expose sensitive data from private sessions without any in-skill disclosure, scoping, or minimization controls.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.