ClawHub
Back to skill

Security audit

meta-ads-creative-builder

Security checks for vulnerabilities and agentic risk

Overview

The bundled workflow uses local files and leaves publishing to you, but its script checks text lengths rather than advertising-policy compliance.

Review copy, factual claims, and policy compliance manually; the report is not an approval gate. Use a dedicated output directory because an existing report is overwritten. The optional unsigned Windows app is outside this assessment; the [publisher release](https://github.com/rapidworker1/Multiple-AI-models-Toolkit/releases/tag/Main) describes a separate general-purpose AI application.

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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The character-count checks match the declared limits, and local file input/output supports the task. However, the declared primary function—creative generation—is absent. Over-limit fields are flagged rather than revised, and no logic checks testimonials, competitor copying, personal attributes, or other policy concerns. The supplied code is a field-completeness and length-reporting utility rather than the advertised copy generator with applied policy checks.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The advertised copy generator with applied policy checks only reports field lengths and prints policy reminders.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The skill advertises applied policy checks, but this script only reports field lengths and prints policy reminders; it does not evaluate whether the supplied copy violates those policies. In an advertising workflow that relies on this report as a validation gate, prohibited or misleading copy could proceed unchecked, creating compliance and account-enforcement risks rather than direct system compromise.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.