ClawHub
Back to skill

Security audit

ab-testing

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only A/B testing guidance skill with some broad activation language but no hidden execution, persistence, or credential handling.

Installers should expect this skill to guide experiment design and may have it read local product-marketing context files when present. Use appropriately collected analytics data for segment analysis and treat its advice as statistical planning guidance, not legal or privacy compliance advice.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
97% confidence
Finding
The trigger description is extremely broad and includes many generic terms like "experiment," "hypothesis," and "which version is better," which can cause this skill to be invoked for conversations that are only loosely related to A/B testing. Over-invocation can misroute user requests, suppress more appropriate specialized skills, and lead to incorrect guidance or context handling across unrelated tasks.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The eval explicitly expects the skill to trigger on casual phrasing ('we want to test like 4 different CTA button colors... is that a good idea?'), which broadens activation boundaries and can cause the A/B testing skill to engage for loosely related comparison or design questions. Over-broad routing increases the chance of misclassification, incorrect skill invocation, and bypass of more appropriate specialist skills, especially in multi-skill agent systems.

Missing User Warnings

Low
Confidence
81% confidence
Finding
This markdown template instructs users to analyze segments such as mobile vs. desktop, new vs. returning visitors, and traffic source, which implies collection and use of behavioral/user-segmentation data. The file does not include any warning or note about privacy considerations, consent, or appropriate handling of user data when performing these analyses.

Static analysis

No suspicious patterns detected.