Security audit

检验报告解读

Security checks for vulnerabilities and agentic risk

Overview

This is a simple lab-report interpretation skill with no executable behavior, network access, credential handling, or persistence, but users should treat its medical ranges as general guidance only.

Install only if you want a Chinese-oriented lab-report interpretation aid. Do not use its built-in ranges as authoritative medical advice; compare results with the reporting lab's reference intervals and units, and seek professional care for abnormal, urgent, or symptom-linked results.

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 (5)

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill provides specific reference ranges and critical values but does not clearly warn that these vary by laboratory method, patient sex, age, pregnancy status, units, and clinical setting. In a medical context, users may over-rely on these values as authoritative, creating a real risk of incorrect interpretation, missed urgency, or false reassurance.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The trigger phrases are broad clinical terms such as '检验报告' and '血糖' that can easily appear in ordinary medical conversation, increasing the chance of unintended activation. In a clinical-assistance context, accidental invocation can cause the model to provide reference-range-driven interpretation when the user did not explicitly request this skill, which may mislead or overstep intended use.

Natural-Language Policy Violations

Medium

Confidence: 91% confidence
Finding: The manifest presents the skill name only in Chinese ("检验报告解读") while the rest of the description is in English, indicating a language-specific user-facing label without any stated language choice or opt-in. This can violate language/locale policy when users are not explicitly offered a preferred language option.

Natural-Language Policy Violations

Low

Confidence: 85% confidence
Finding: Most labels, critical-value descriptions, and all trigger phrases are written only in Chinese, which effectively constrains use to a specific language without stating that limitation or offering opt-in. The policy requires avoiding forced language or locale choices unless they are explicitly documented and justified.

Static analysis

No suspicious patterns detected.