Security audit

Deeptrack

Security checks across malware telemetry and agentic risk

Overview

This paid plugin is presented as a deep research agent, but the code only returns research templates after payment and sends payment tokens to an unsecured HTTP gateway.

Review carefully before installing or paying. Treat this as a paid template generator, not a working deep research service, and avoid submitting payment transaction IDs to the default HTTP gateway unless the publisher provides clear HTTPS verification and accurately describes what users receive.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (5)

Intent-Code Divergence

High

Confidence: 98% confidence
Finding: The tool description promises 'deep web research' and 'multi-source intelligence synthesis,' but after payment verification the implementation only returns a static prompt/framework. This is deceptive monetization behavior that can cause users to pay for capabilities the skill does not actually provide, undermining trust and potentially enabling fraud-like outcomes.

Intent-Code Divergence

High

Confidence: 98% confidence
Finding: The competitor-analysis tool advertises in-depth analysis, matrices, moat evaluation, and SWOT output, but the code only emits a static framework template. Charging for or presenting this as completed analysis is misleading and can cause users to rely on nonexistent analytical work.

Intent-Code Divergence

High

Confidence: 98% confidence
Finding: The trend-forecasting tool claims signal detection, scenario planning, and timeline projection as if it performs forecasting, but it only returns a generic forecasting framework. This mismatch is dangerous because users may pay for or act on output they believe is generated analysis rather than a blank template.

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: The help text says users will get 'expert research frameworks and analysis,' but the implementation only returns frameworks. While less severe than the tool descriptions themselves, this still contributes to misleading claims around a paid feature.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The payment token is sent to a remote gateway, and the default gateway uses plain HTTP rather than HTTPS. This exposes sensitive transaction identifiers to interception or tampering in transit, and users are not clearly warned that their token will be disclosed to a third-party network endpoint.

VirusTotal

60/60 vendors flagged this plugin as clean.

View on VirusTotal