ClawHub
Back to skill

Security audit

Ads Compliance Review

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only advertising compliance guidance skill with some overly broad marketing-optimization scope, but no hidden execution, persistence, credential use, or destructive behavior.

Installers should treat this as a combined ad compliance and marketing-optimization advisor, not a narrowly scoped compliance checker. Avoid giving it credentials or account access unless separately authorized, and review any budget, bidding, or campaign-execution recommendations before acting on them.

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
  • 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 (4)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger phrases are broad enough to match generic marketing and growth requests, not just compliance-review tasks. Overbroad routing logic increases the probability that unrelated user requests are captured by this skill, causing policy-review tooling to act as a general ad strategy advisor and potentially bypass intended orchestration boundaries.

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The skill is presented as an ads compliance reviewer, but its trigger scope expands into campaign execution, revenue growth, budget optimization, and operational ad-running tasks. This creates scope drift that can cause the agent to invoke the skill for broad marketing execution requests, potentially bypassing narrower controls or human review expected for a compliance-only function.

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The output contract and workflow promise execution-ready action plans and KPI-driven optimization, which materially exceed compliance review. In an agentic environment, this can turn a supposedly advisory safety screen into a general growth optimizer, increasing the chance of unsafe autonomy, misrouting, or unauthorized campaign recommendations under the guise of compliance review.

Description-Behavior Mismatch

Medium
Confidence
91% confidence
Finding
The examples and guidance include budget splits, bidding adjustments, test matrices, and expected KPI movement, signaling that the skill will provide tactical media-buying advice beyond compliance review. That broadens the operational authority implied by the skill and makes accidental invocation for non-compliance tasks more likely, undermining least-privilege skill design.

Static analysis

No suspicious patterns detected.