Security audit

Skill Guangjiao

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Canton Fair lead-discovery and outreach-assistance workflow, with privacy and anti-spam caveats users should still take seriously.

Install only if you intend to use Yunlv/TradeGPT for Canton Fair prospecting. Before using exported contacts or generated outreach, confirm the data source is lawful for your market, avoid bulk unsolicited messaging, keep the TRADEGPT_API_KEY protected, and delete stored lead exports when no longer needed.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

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 (9)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The trigger list contains broad terms such as “参展商”, “展会数据”, and “Canton Fair” that can appear in ordinary business conversations unrelated to this specific skill. Over-broad activation can cause unintended invocation of a lead-mining workflow that handles contact data and generates outreach content, increasing the chance of inappropriate data processing or user confusion about why the skill was used.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 86% confidence
Finding: The skill advertises extracting, exporting, and locally storing exhibitor contact information, but it does not present a prominent pre-use notice explaining retention, storage location, third-party API transfer, and consent/compliance implications before the workflow begins. In a lead-generation context, this increases privacy and compliance risk because users may trigger processing of personal/business contact data without informed awareness of how the data will be handled.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The manifest explicitly advertises exhibitor contact discovery plus email and WhatsApp message generation, but provides no privacy, consent, or anti-spam warning. In this context, that omission can facilitate scraping, unsolicited outreach, or misuse of personal/business contact data, especially because the skill is designed for lead mining and outbound messaging.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 79% confidence
Finding: The manifest mixes English with Chinese-only trigger phrases and descriptions for core usage, but it does not state that the skill is intentionally limited to Chinese-speaking users or offer alternative language options. Under the policy, locale-specific behavior should be opt-in or clearly justified when a specific language is effectively required.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: This manifest defines several activation triggers such as "展会客户", "参展商", "展会数据", and "展品分类" that are broad and not clearly limited to Canton Fair-specific requests. Although some entries are specific, the mixed broad terms can cause unintended invocation for general exhibition or lead-generation conversations because no exclusion conditions or negative examples are provided.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The document explicitly instructs staff to collect and use personal contact details such as phone numbers and email addresses, then immediately contact leads over WhatsApp, WeChat, and email, but provides no notice about consent, lawful basis, retention, or privacy handling. In a lead-mining skill focused on trade-show prospecting, this increases the risk of noncompliant personal-data processing, unwanted outreach, and potential regulatory or reputational harm.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The template explicitly encourages outreach based on a prospect's import history and named competitor relationship, but provides no guardrails about whether that data was lawfully obtained, whether it contains personal data, or whether its use complies with privacy, marketing, and trade-data restrictions. In a lead-mining skill focused on Canton Fair and exhibitor discovery, this omission increases the risk of misuse for non-compliant profiling, intrusive cold outreach, or improper use of third-party commercial intelligence.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: This markdown file contains user-facing natural language content exclusively in Chinese, while also mixing in English labels, but it does not explicitly offer users a language or locale choice. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale limitation is clearly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 75% confidence
Finding: A language or locale policy issue can arise when a skill implicitly forces one language for all users without opt-in. This file is fully Chinese-language content and does not indicate that the user can choose another language or that the skill is intended only for a Chinese-language audience.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.