Google Ads Strategy: Search Campaigns, Keywords, Ad Copy, Negative Keywords, Quality Score

• Outcome: Review
• Latest audit: Sep 16, 2026 · 5:54 AM UTC
• Version: 1.0.0

Overview

The skill is mostly normal Google Ads guidance, but it defaults to local AdKit CLI execution for campaign uploads, creation, and publishing without clearly requiring user confirmation for high-impact ad-account changes.

Review before installing if you do not want a skill that may use a local AdKit CLI for Google Ads operations. Require explicit approval before any keyword upload, campaign creation, ad publishing, budget change, or customer-data sharing such as Enhanced Conversions.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (8)

Whitespace Padding

Medium

• Category: Prompt Injection

Content

## Campaign Types — B2B SaaS

| Campaign       | What it targets                    | Notes                                                                                     |
| -------------- | ---------------------------------- | ----------------------------------------------------------------------------------------- |
| **Brand**      | Your brand name + variants         | Bid even if you rank organically — competitors will otherwise serve above your result. Cheapest click in the account. |
| **Competitor** | Named competitor terms             | Requires comparison landing pages. Separate budget and bidding from solution campaigns.   |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Whitespace Padding

Medium

• Category: Prompt Injection

Content

## Match Type Decision Table

| Scenario                                                  | Match type                                  | Why                                                                                      |
| --------------------------------------------------------- | ------------------------------------------- | ---------------------------------------------------------------------------------------- |
| Budget under $10K/month                                   | Exact + Phrase only                         | Not enough data for broad match to learn. Every wasted dollar hurts.                    |
| Under 30 conversions/month                                | Exact + Phrase only                         | Broad match needs 30+ conversions/month minimum to optimize.                            |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Whitespace Padding

Medium

• Category: Prompt Injection

Content

Add all of these before launching. Google uses them to fill ad space and improve CTR.

| Asset               | Guidance                                                                                                                    |
| ------------------- | --------------------------------------------------------------------------------------------------------------------------- |
| **Sitelinks**       | Add beyond the display cap (6 desktop / 8 mobile) — Google selects the most relevant. Exception: limit to 2 for max-promotion pinning strategy. |
| **Callouts**        | ≤ 25 chars each, not clickable. Specific and credible ("SOC 2 Type II", "14-day free trial"). Not vague fluff ("Best in class"). |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Natural-Language Policy Violations

Medium

• Confidence: 89% confidence
• Finding: Lines L069-L073 instruct users to add language terms as negative keywords to block searches from certain languages. This imposes a language-based exclusion policy in natural-language guidance without offering user choice or documenting a region-specific compliance/business justification, which fits the language/locale policy violation criteria.

Shadow Command Trigger

Medium

• Category: Trigger Abuse
• Confidence: 70% confidence
• Finding: Shadow Command Trigger: 'search ads' conflicts with built-in command 'search'

Description-Behavior Mismatch

Medium

• Confidence: 92% confidence
• Finding: The manifest and description frame the skill as planning and optimization guidance, but this section routes the agent into local AdKit-based execution and campaign operations. That mismatch can mislead users and orchestration systems about the skill's actual capabilities, increasing the chance of unexpected command execution or external side effects.

Context-Inappropriate Capability

Medium

• Confidence: 95% confidence
• Finding: The skill is presented as advisory guidance for Google Ads strategy, but it instructs the agent to execute a local command (`adkit status`) to probe the host environment. That expands behavior from analysis into system interaction without clear user authorization or necessity, creating unnecessary exposure to local-environment access and tool-execution risks.

Missing User Warnings

Low

• Confidence: 83% confidence
• Finding: This markdown skill directs users to use competitor reviews and competitor landing pages as inputs, which can involve collecting and processing third-party content. The document provides no user-facing warning or caveat about responsible handling, platform restrictions, or privacy implications, despite describing data-gathering behavior.

Static analysis

No suspicious patterns detected.