Security audit

大厂永生

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill gives workplace communication advice, with the main caution that users should redact sensitive company or personal details before sharing conversations.

Before installing, understand that the skill is meant for analyzing workplace conflict messages and drafting replies. Do not paste raw company-confidential conversations, customer data, personal identifiers, credentials, or private message metadata unless necessary; redact or summarize sensitive details first.

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 (3)

Missing User Warnings

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The skill explicitly encourages users to paste chat screenshots, call transcripts, emails, and enterprise messaging records, but provides no privacy warning, minimization guidance, or redaction requirement. Because these materials often contain personal data, confidential business information, internal discussions, and third-party data, the skill creates a strong risk of unnecessary sensitive-data exposure.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The entire skill is authored as Chinese-only and all example prompts, output expectations, and user-facing language assume Chinese, but there is no statement that the skill is limited to Chinese users or that users may choose another language. This creates a natural-language locale policy concern because the skill implicitly enforces a single language without opt-in or justification.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger list includes broad everyday phrases such as '帮我看看这个', '分析一下这段对话', and common workplace terms that can appear in many unrelated conversations. This can cause unintended activation on general chat-analysis requests, increasing the chance the skill processes sensitive workplace content when the user did not explicitly intend to invoke this specialized behavior.
Content: No source excerpt is available for this finding.