ClawHub
Back to skill

Security audit

Lightcone Session

Security checks for vulnerabilities and agentic risk

Overview

This skill is mainly a disclosed remote browser/desktop control helper, but it also exposes cloud-computer shell commands and credential/session handling without enough safety scoping.

Review before installing. Use it only with accounts and sites you are authorized to control, avoid entering real secrets unless necessary, do not capture or share sensitive screenshots or page HTML, and treat the debug command as powerful shell access inside the cloud computer rather than ordinary browser automation.

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
  • 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 (3)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The documented debug action allows arbitrary shell command execution on the cloud computer, which materially exceeds the skill's stated purpose of step-by-step browser/desktop control. Even if intended for troubleshooting, this creates a powerful execution primitive that could be used to run reconnaissance, download payloads, access local session artifacts, or pivot beyond normal UI automation boundaries.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill documents shell-command execution as a normal action without any cautionary language about the risks of running commands in the cloud computer. This omission can normalize dangerous use, leading users or downstream agents to execute commands without understanding that they may alter the environment, expose data, or perform unintended network/system operations.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The example explicitly walks through entering login credentials and extracting page HTML, but provides no warning about sensitive data handling, consent, session security, or secrets exposure in screenshots and extracted content. In a remote browser/session tool, this increases the likelihood of unsafe handling of passwords, authenticated content, and personal or business data.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.