Security audit

审计追踪

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only finance audit guidance skill with disclosed sensitive-data use; the main caution is broad activation triggers, not hidden execution.

Use this skill for finance audit assistance only with data you are authorized to process. Because it may activate on broad finance/audit terms, review outputs carefully, limit shared datasets to what is necessary, and treat suggested freezes, collections, tax actions, and risk labels as requiring human professional review.

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

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

Findings (3)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The top-level description says the skill triggers on many broad finance and audit keywords, which can overlap with ordinary conversation or general advisory requests. This can cause unintended activation and route sensitive financial discussions into a specialized workflow the user did not explicitly request, increasing the chance of over-collection, irrelevant guidance, or disclosure of audit-oriented outputs.

Natural-Language Policy Violations

Medium

Confidence: 88% confidence
Finding: L003 的自然语言描述整体仅以中文呈现，并未说明是否支持其他语言或允许用户选择输出语言。按该技能说明，存在默认强制单一语言的倾向，可能不符合要求提供语言/locale 选择的组织政策。

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The manifest trigger list includes very broad terms like 'audit', '异常交易', '风险检测', '应收账款', and '发票核验' without contextual constraints. In a finance environment, these terms are common and may invoke the skill during routine discussion, causing unintended processing of potentially sensitive accounting or tax data and creating opportunity for prompt-routing abuse or user confusion.

Static analysis

No suspicious patterns detected.