ClawHub
Back to skill

Security audit

VeriGlow Agent Map

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed website data-extraction helper, but it is too broadly scoped around hidden APIs and automation without clear user authorization or privacy limits.

Install only if you intend to use VeriGlow for public, authorized web data extraction. Do not send private, authenticated, confidential, or access-controlled URLs to the service, and do not use hidden API or automation recipes to bypass site rules, authentication, rate limits, or geographic restrictions.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The description omits clear warnings that querying arbitrary URLs and following browser automation or hidden-API recipes may transmit sensitive targets to a third-party service, interact with websites in ways the user did not intend, or facilitate scraping of restricted resources. This is more dangerous here because the skill markets discovery of internal APIs and automation steps, yet provides no upfront privacy, authorization, or operational safety boundaries.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill is framed for generic web data extraction, hidden API discovery, and browser automation across arbitrary sites, which can cause an agent to invoke it for many loosely related requests without sufficient trust, authorization, or sensitivity checks. In context, this is risky because the skill explicitly encourages use against third-party URLs and hidden/internal APIs, increasing the chance of privacy violations, policy bypass attempts, or unintended interaction with sensitive sites.

Static analysis

No suspicious patterns detected.