Marketing Skills

• Outcome: Review
• Latest audit: Jul 10, 2026 · 2:39 AM UTC
• Version: 0.1.2

Overview

This marketing skill appears mostly educational, but it includes under-scoped guidance for live ad/social account actions and large-scale data collection that users should review before installing.

Review this skill before installing, especially if your agent has access to ad accounts, social publishing tools, analytics, CRM/email systems, or scraping services. Keep live account actions draft-only unless a human confirms the exact change, audience, platform, timing, and spend; require consent and privacy review for tracking, segmentation, lead capture, and behavioral emails; and verify platform terms before any bulk data collection.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (21)

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The skill description contains broad trigger phrases such as "test this change" and "hypothesis," which can match ordinary product, engineering, or even general discussion outside intentional A/B testing requests. This can cause inappropriate skill activation, leading the agent to steer conversations into experimentation workflows when the user did not ask for that, creating misrouting and potentially lower-quality or misleading assistance.

Missing User Warnings

Medium

• Confidence: 86% confidence
• Finding: The skill gives detailed guidance for implementing analytics, identifiers, campaign parameters, and event properties, but it does not prominently require a privacy review or user-facing warning before suggesting collection patterns. Although the document later mentions consent and avoiding PII, those safeguards are secondary and easy to miss, which can lead users to deploy tracking that exceeds consent, collects unnecessary data, or violates regional privacy rules.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The manifest description contains very broad trigger phrases such as 'competitor comparison,' 'comparison page,' and product-vs-product patterns, which can cause the skill to activate on generic requests without tight scoping. Over-broad activation increases the chance the wrong skill handles a request, leading to irrelevant guidance, unexpected prompt injection surface expansion, or unintended use in contexts not specifically about creating SEO competitor pages.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The skill description uses broad trigger phrases like 'edit this copy,' 'proofread,' and 'make this better,' which can match many ordinary user requests and cause the agent to invoke this skill outside its intended marketing-copy scope. Over-broad activation can misroute user requests, override more appropriate skills, and increase the chance that unrelated content is processed under persuasive marketing-editing heuristics.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill explicitly recommends behavior-based and usage-based targeting such as activity level, feature use, inactivity, and other engagement signals, but it provides no guardrails around consent, lawful basis, minimization, or sensitivity of the underlying telemetry. That omission can lead downstream agents or operators to build campaigns that profile users or trigger outreach from potentially sensitive product behavior in ways that violate privacy expectations or regulatory requirements.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill recommends personalized usage, billing, and account-status emails using metrics, plan details, payment state, and renewal information without warning that these data elements are sensitive and should be handled carefully. If used naively, this can expose private account details in email, encourage overcollection or overpersonalization, and create compliance and trust risks, especially where emails are visible to unintended recipients or sent without appropriate notice.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The activation description includes many broad trigger phrases such as "calculator," "generator," "interactive tool," and "free resource," which can match ordinary user requests outside the intended scope. Over-broad activation can cause the agent to invoke this marketing-oriented skill in unrelated contexts, steering conversations toward lead-generation tactics or promotional framing without the user explicitly asking for that.

Missing User Warnings

Medium

• Confidence: 87% confidence
• Finding: The skill gives detailed advice on gating tools behind email capture and using nurture or follow-up marketing sequences, but it does not instruct the agent to warn about consent, privacy obligations, or transparent disclosure of how collected addresses will be used. In practice, this can lead the agent to help design deceptive or non-compliant lead capture flows that collect personal data without adequate notice or permission.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The description includes many generic trigger terms such as 'launch,' 'announcement,' and 'product update,' which can cause the skill to activate for broad business conversations that do not actually need this specific skill. Over-broad activation can misroute user requests, crowd out more appropriate skills, and increase the chance of irrelevant or lower-quality guidance being returned.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The trigger description is very broad and matches generic phrases like 'marketing ideas,' 'how to market,' and 'ways to promote,' which can cause the skill to activate for a large range of ordinary business conversations. Overbroad activation increases the chance the agent will inject this skill in contexts where it is not specifically intended, including situations involving sensitive growth tactics or compliance-relevant advice, reducing user control and making downstream risky suggestions more likely to surface.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The skill includes tactics that can affect privacy, compliance, or platform integrity—such as pixel sharing, remarketing audience sharing, competitor-term ads, parasite SEO, engagement pods, giveaways, and scanning/security-checker ideas—without any warnings or guardrails. In this context, omission of cautions is dangerous because the skill presents these tactics as 'proven' options, which may lead an agent to recommend actions that violate privacy laws, ad platform rules, anti-spam policies, or ethical boundaries without surfacing risks.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The trigger description is very broad, matching many generic terms such as 'psychology,' 'persuasion,' and 'decision-making,' which can cause the skill to activate in contexts far beyond marketing. This can lead to unintended routing, inappropriate advice, or overshadowing more suitable skills, especially when sensitive persuasion-related topics are discussed.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The skill description includes broad activation phrases such as generic requests to 'optimize' or 'improve conversions' on many page types, which can cause the agent to invoke this skill in situations that are only loosely related. Overbroad routing is dangerous because it can overshadow more specific skills, produce irrelevant guidance, and increase the chance that the agent applies the wrong workflow to user input.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The manifest description contains many broad trigger phrases such as 'ad copy,' 'audience targeting,' and 'ad campaign,' which are common terms that may appear in general marketing discussions. This can cause the skill to activate outside its intended scope and steer the agent into ad-platform-specific guidance or account-affecting workflows without strong user intent, increasing the chance of inappropriate tool use or unsafe delegation.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The skill explicitly states it has 'direct access to ad platform accounts' but does not require a user-facing warning, confirmation, or clear separation between analysis and account-changing actions. In context, this is more dangerous because paid ads actions can immediately spend money, alter targeting, launch creatives, or disrupt business campaigns, so ambiguous authority can lead to unauthorized or accidental financial and operational impact.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The activation description is very broad and triggers on common terms like 'pricing' and 'free trial' without meaningful scope limits, so the agent may invoke this skill in many loosely related conversations. That can cause inappropriate skill routing, overshadow more suitable skills, and lead to advice being given under the wrong context, which is a real security/safety issue in agent behavior even though the content itself is not overtly malicious.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The activation description contains very broad trigger phrases such as 'pages at scale,' 'directory pages,' 'location pages,' and '[keyword] + [city] pages,' which can match common user requests outside a narrowly intended programmatic SEO context. This can cause unintended skill invocation, steering the agent into SEO-specific guidance when the user may have meant content architecture, web development, or other unrelated tasks, increasing the risk of incorrect or over-scoped behavior.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The trigger description is very broad and includes generic marketing terms like 'affiliate,' 'ambassador,' and 'word of mouth,' which could cause this skill to activate for loosely related requests. That can route users into the wrong skill, producing irrelevant guidance or exposing unnecessary business-analysis behavior when another specialized skill would be more appropriate.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The manifest description includes very broad trigger phrases such as 'audit, review, or diagnose SEO issues on their site,' which can cause the skill to activate for generic troubleshooting or website-review requests beyond a narrowly intended SEO context. Over-broad activation increases the chance of incorrect routing, unintended tool use, or the skill taking over conversations where a more appropriate skill should handle the request.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill claims 'direct access to a scheduling platform that publishes to all major social networks' but does not require explicit confirmation, preview, or safety checks before account-affecting actions. In an agentic environment, this can lead to unintended posting, scheduling, or brand-impacting actions on linked social accounts from ambiguous or incomplete user requests.

Unrestricted Tool Access

Medium

• Category: Excessive Agency

Content

Gather 500-1000+ posts from your identified creators for analysis:

**Tools:**
- **Apify** — LinkedIn scraper, Twitter scraper actors
- **Phantom Buster** — Multi-platform automation
- **Export tools** — Platform-specific export features

• Confidence: 95% confidence
• Finding: Tools:*

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.