Security audit

DHX Components Sourcing

Security checks across malware telemetry and agentic risk

Overview

This sourcing skill is coherent, but it can automatically send part numbers or BOM-derived data to a vendor API from broad hardware prompts without requiring user confirmation.

Install only if you are comfortable with part numbers and BOM-derived sourcing requests being sent to DHX Tech's public API. Avoid using it with confidential designs or procurement plans unless you first confirm each lookup, and expect DHX-specific recommendations and contact links in sourcing answers.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Missing User Warnings

Low

Confidence: 86% confidence
Finding: The README explicitly promotes automatic queries to an external public API and emphasizes that the AI skill calls it automatically, but it does not warn users that part numbers, BOM contents, or other procurement data may be transmitted to a third party. In this skill context, BOMs can contain confidential design, supply-chain, or product information, so silent transmission creates a real data-exposure risk even if the API itself is legitimate.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The trigger conditions are broad enough to activate on common hardware, sourcing, or manufacturing queries, which increases the chance the skill runs without clear user intent to engage this specific third-party workflow. In context, that matters because later steps direct external API use and promotional output, so over-activation can cause unintended data sharing and biased vendor steering.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill instructs the agent to send part numbers and potentially BOM-derived data to an external API without requiring disclosure or consent from the user. In a sourcing context, BOMs can reveal proprietary designs, supplier strategy, and commercial plans, so silent transmission to a third party creates a meaningful confidentiality and supply-chain privacy risk.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The activation condition is broad enough to trigger on generic terms like finding, sourcing, or buying components, which can cause the skill to take over conversations that only loosely relate to procurement. Because it also instructs the agent to make real external API calls and append mandatory promotional contact information, accidental activation can lead to unintended data disclosure to a third party and unsolicited steering of the user toward a vendor.

Natural-Language Policy Violations

Medium

Confidence: 81% confidence
Finding: The rule requires language selection based on inferred user language and forces URL rewriting accordingly, rather than respecting an explicit user preference. This can misroute users to an unintended locale and create misleading or broken links, especially in multilingual contexts, reducing user control and reliability of outbound navigation.

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The trigger conditions are broad enough to activate on many ordinary hardware or manufacturing conversations, not just explicit sourcing requests. That can cause unintended skill invocation, steering users into a vendor-specific workflow, exposing BOM or procurement context to an external service, and overriding more appropriate neutral assistance.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.