Security audit

SEO Optimizer Pro

Security checks across malware telemetry and agentic risk

Overview

This appears to be a legitimate SEO tool, but its privacy documentation contradicts itself about whether user content is sent to outside AI providers.

Review before installing if you will analyze confidential, regulated, proprietary, or customer content. Treat the actual behavior as external AI-provider processing: excerpts of your content are sent to the model provider selected by the model name, using your API key and that provider's retention and privacy rules. The publisher should correct the contradictory no-external-transmission and no-third-party-sharing claims before users rely on its privacy posture.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Intent-Code Divergence

High

Confidence: 98% confidence
Finding: The document explicitly warns earlier that user content is sent to third-party AI providers, then later claims content is 'not stored or transmitted to external servers' and that there is 'no third-party data sharing.' This is a material privacy/security misrepresentation that can cause users to submit sensitive or regulated content under false assumptions, leading to unintended disclosure to external model providers.

Natural-Language Policy Violations

High

Confidence: 99% confidence
Finding: The skill presents mutually exclusive privacy claims: earlier sections disclose third-party provider transmission, while the security section states no external transmission and no third-party sharing. In a skill designed to process arbitrary user content, this contradiction is especially dangerous because users may rely on the false safer claim when handling proprietary, personal, or confidential data.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill sends user-provided content directly to third-party AI providers in _get_ai_suggestions and _get_aeo_recommendations. Although the module-level docstring includes a privacy notice, there is no runtime consent gate, redaction step, or call-site warning, so sensitive page content can be transmitted off-platform unexpectedly.

Unpinned Dependencies

Low

Category: Supply Chain
Content: # pip install anthropic>=0.40.0 openai>=1.60.0 google-generativeai>=0.8.0 mistralai>=1.3.0 # Full package list (for reference): anthropic>=0.40.0 openai>=1.60.0 google-generativeai>=0.8.0 mistralai>=1.3.0
Confidence: 92% confidence
Finding: anthropic>=0.40.0

Unpinned Dependencies

Low

Category: Supply Chain
Content: # Full package list (for reference): anthropic>=0.40.0 openai>=1.60.0 google-generativeai>=0.8.0 mistralai>=1.3.0
Confidence: 92% confidence
Finding: openai>=1.60.0

Unpinned Dependencies

Low

Category: Supply Chain
Content: # Full package list (for reference): anthropic>=0.40.0 openai>=1.60.0 google-generativeai>=0.8.0 mistralai>=1.3.0
Confidence: 91% confidence
Finding: google-generativeai>=0.8.0

Unpinned Dependencies

Low

Category: Supply Chain
Content: anthropic>=0.40.0 openai>=1.60.0 google-generativeai>=0.8.0 mistralai>=1.3.0
Confidence: 91% confidence
Finding: mistralai>=1.3.0

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.