ClawHub
Back to skill

Security audit

Agentic Test Engineer

Security checks for vulnerabilities and agentic risk

Overview

This is a testing guidance skill with no executable code or install-time behavior, though users should be careful with screenshots, DOM snapshots, and broad test-debugging prompts.

Before installing, treat this as an assistant for generating and reviewing test artifacts, not an automatic authority to change production tests. Avoid sending screenshots, DOM snapshots, CI logs, or coverage files containing secrets, customer data, or regulated data to external visual-AI or SaaS tools unless your organization permits it, and review any generated or self-healed tests before merging.

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
95% confidence
Finding
The skill prominently promotes 'self-heals broken selectors using visual AI' but does not consistently warn, at the point of use, that screenshots, DOM snapshots, or UI metadata may be sent to external tooling or SaaS services. In testing environments, those artifacts can contain secrets, customer data, internal URLs, or regulated information, so incomplete disclosure can lead to unintentional data exfiltration or policy violations.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger set includes very broad phrases such as 'why is my test failing' and 'run E2E tests for [workflow]', which can overlap with ordinary debugging or support requests and cause the skill to activate in contexts where the user did not intend autonomous test-generation or self-healing behavior. In a skill that can recommend selector changes, coverage actions, and CI-oriented workflows, over-broad activation increases the chance of unintended guidance, excessive data exposure to the skill context, or unsafe automation suggestions.

Static analysis

No suspicious patterns detected.