ClawHub
Back to skill

Security audit

Ad Creative Analysis

Security checks for vulnerabilities and agentic risk

Overview

This skill reads user-provided ad creative folders to produce marketing analysis and does not show hidden execution, persistence, credential access, or data exfiltration behavior.

Installers should understand that the skill will read filenames, ad media, transcripts, and optional metadata from any directory they provide. Use it on folders intended for ad review, and avoid pointing it at broad personal or business directories unrelated to the analysis task.

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 (2)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill's trigger phrases are broad enough to overlap with ordinary marketing-analysis requests, which can cause the agent to invoke this skill in contexts the user did not specifically intend. That creates scope ambiguity and can lead to unintended access to local directories, media files, or metadata during routine requests, especially in systems where skills auto-activate from natural language.

Natural-Language Policy Violations

Low
Confidence
83% confidence
Finding
This markdown reference repeatedly provides hook and CTA examples only in English and presents them as default patterns for analysis output. Because SQP-3 applies to language/locale policy violations, this can be read as forcing a specific language without any documented user choice or locale justification.

Static analysis

No suspicious patterns detected.