Security audit

Cabinet Kitchen Marketing Kit

Security checks for vulnerabilities and agentic risk

Overview

This is a content-only marketing prompt kit with disclosed compliance guardrails and no executable behavior, though users should still verify generated advertising claims before publication.

Install only if you want a Nevada/Southwest-focused cabinet marketing prompt kit. Before publishing generated ads, SEO pages, review requests, or referral letters, verify license numbers, EPA RRP status, warranty terms, financing disclosures, and environmental certifications, and remove any unsupported 'lifetime' or 'eco-friendly' wording.

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 (5)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: This markdown file describes the skill’s purpose and outputs but does not define clear trigger phrases, scope boundaries, or exclusion conditions for invocation. Under SQP-1 for markdown files, that absence of specificity can lead to unintended activation for loosely related reputation or referral tasks.

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The prompt claims that all outputs must enforce hard compliance blocks, but later sections include content that can generate environmental marketing claims without consistently requiring substantiation. This creates a policy contradiction that may cause the agent to emit non-compliant ad copy or SEO terms, exposing users to deceptive advertising or regulatory risk.

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The hard blocks prohibit 'lifetime guarantee' language unless narrow conditions are met, yet the template later includes a 'Lifetime HW' headline pattern. An agent following the template may produce an unsubstantiated lifetime warranty claim in ad copy, creating consumer protection and false advertising exposure.

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The keyword clusters include an unconditional 'eco friendly kitchen remodel' term even though the prompt says hard blocks apply to keywords and restricts environmental claims without documentation. This inconsistency can directly cause generation of misleading SEO content and ad-adjacent marketing claims lacking substantiation.

Natural-Language Policy Violations

Low

Confidence: 82% confidence
Finding: The service area and all prescribed ad/SEO outputs are framed entirely for English-language marketing, and the prompt does not offer any user choice for generating Spanish or other locale variants. For a marketing skill, this can amount to a language/locale policy issue when the skill implicitly fixes one language without opt-in.

Static analysis

No suspicious patterns detected.