All-in-one B2B lead generation toolkit aggregating customs trade intelligence, global company deep due-diligence and LinkedIn professional network data into asingle workflow. Analyze HS code market distribution, trade trends and company trade-share rankings to size any product market. Profile a singlecompany's real trade scale, partners, products and ports from 220+ countries of customs records. Pull macro country-level trade overviews, top buyersand suppliers and US import statistics. Run deep company background checks covering employees, shareholders, ultimate beneficial owners (UBO) anddecision-makers, and map LinkedIn professional networks including colleagues, alumni, career histories and education backgrounds. Built for exporters,sourcing agents, sales teams and B2B prospecting specialists to discover overseas buyers, validate suppliers and accelerate cross-border customeracquisition through combined customs-data, corporate-database and LinkedIn intelligence.

• Outcome: Review
• Latest audit: Aug 4, 2026 · 2:37 AM UTC
• Version: 1.0.0

Overview

This skill appears to be a real paid B2B lead-generation integration, but it needs review because it enables broad person-level contact and LinkedIn-style profiling without clear privacy or lawful-use limits.

Review this before installing if you will use it for people, employee, LinkedIn, or contact-data searches. Only run lookups when you have a lawful business purpose, expect paid API charges, protect ~/.upkuajing/.env, and periodically delete task_data results that contain personal or business data.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Rogue Agent: Self-Modification, Session Persistence
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger

Findings (12)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 97% confidence
• Finding: The skill instructs the agent to read environment variables, read and write files in the user's home directory, and make networked API calls, but it does not declare corresponding permissions. This creates a transparency and control gap: users and the platform may not realize the skill can access local secrets and persist data, increasing the chance of unintended secret exposure or unauthorized file modification.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The trigger list is extremely broad and overlaps with many ordinary business, recruiting, sourcing, and research requests, so the skill may be invoked in situations where the user did not clearly intend to perform paid lead generation or sensitive person/company lookups. Overbroad routing can cause unnecessary data collection, external API calls, and privacy-impacting actions without sufficiently specific user intent.

Missing User Warnings

High

• Confidence: 95% confidence
• Finding: The skill is designed to collect and correlate potentially sensitive data about employees, shareholders, UBOs, decision-makers, colleagues, alumni, education, and career history, yet it provides no upfront privacy warning, lawful-use notice, or data-handling constraints. In this context, the aggregate cross-source enrichment makes profiling and deanonymization more powerful, raising misuse risk for surveillance, targeting, or intrusive due diligence.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The document explicitly exposes filters and returned fields related to employee and core-member contact data, including phone, email, WhatsApp, and a sample employee contact display field, without any privacy, legal-use, consent, or access-control guidance. In a lead-generation and due-diligence skill, this materially increases the risk of misuse for scraping, profiling, or unsolicited outreach against identifiable individuals.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The API documentation explicitly supports searching person records using contact-data presence filters such as phone, WhatsApp, email, website, and social accounts, but provides no privacy, legal-use, consent, or data-handling constraints. In the context of a B2B lead-generation and due-diligence skill, this materially increases the risk of targeted profiling, mass contact harvesting, and misuse of personal data at scale.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The API reference explicitly documents filters and result fields for employee and core-member contact availability, including phone, email, WhatsApp, social accounts, and a displayed employee contact. In a B2B lead-generation and LinkedIn-networking skill, this enables targeted discovery and potential bulk harvesting of personal or semi-personal contact data without any stated privacy constraints, lawful-basis guidance, consent requirements, or anti-abuse limits, increasing the risk of doxxing, spam, spearphishing, and privacy-law violations.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The document explicitly exposes filters and result fields for discovering whether individuals have phone numbers, email addresses, WhatsApp, websites, and social accounts, but provides no warning, consent requirements, or usage restrictions for privacy-sensitive personal data. In the context of a lead-generation and LinkedIn network-mapping skill, this materially increases the risk of targeted scraping, profiling, and outreach against identifiable people.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The script sends person-search parameters to an external API and persists returned result/task metadata locally, but it provides no explicit user-facing notice or consent step about transmitting potentially sensitive personal/company search data or storing it on disk. In a lead-generation and due-diligence context, this increases privacy, compliance, and data-handling risk because operators may unknowingly process personal data across jurisdictions and leave recoverable local records.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: This script fetches and outputs a company's LinkedIn employee list, which is personal/professional data tied to identifiable individuals, but provides no consent check, purpose limitation, privacy notice, or access control at the script level. In a lead-generation and due-diligence context, this increases the risk of unauthorized profiling, scraping, or downstream misuse of employee data, especially if exposed to broad users or automation.

Credential Access

High

• Category: Privilege Escalation

Content

"envFilePath": str(env_file)
        }

    # 保存到 .env 文件
    try:
        with open(env_file, 'w', encoding='utf-8') as f:
            f.write(f"{API_KEY_ENV}={api_key}\n")

• Confidence: 88% confidence
• Finding: .env

Session Persistence

Medium

• Category: Rogue Agent

Content

### Account Top-up
When API response indicates insufficient balance:
1. Create top-up order (`auth.py --new_rec_order`)
2. Send the payment page URL to the user, guide them to open & pay; proceed after they confirm success.

### Get Account Information

• Confidence: 75% confidence
• Finding: Create top-up order (`auth.py --new_rec_order`) 2. Send the payment page URL to the user, guide them to open & pay; proceed after they confirm success. ### Get Account Information `python scripts/aut

Unpinned Dependencies

Low

• Category: Supply Chain

Content

httpx>=0.23.0

• Confidence: 95% confidence
• Finding: httpx>=0.23.0

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical

• Code: suspicious.exposed_secret_literal
• Location: scripts/common.py:196