ClawHub
Back to skill

Security audit

Supplier Compliance Check（医械供应商合规核查）

Security checks for vulnerabilities and agentic risk

Overview

This document-only supplier compliance checklist has proportionate instructions and no executable payload or privileged access requests.

Use this for medical-device supplier reviews, anonymize confidential supplier information, and have qualified staff verify regulatory conclusions. The bundled checksum does not match SKILL.md; do not rely on its integrity attestation. Referenced external resources and companion skills are outside this package 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 (1)

A generic purchasing-risk question unconditionally triggers the medical-device compliance skill.

Low
Category
Trigger Abuse
Confidence
94% confidence
Finding

Line 69 includes “Purchasing wants the cheap one, how risky is it,” while lines 63 and 72 instruct activation whenever any listed phrase matches. This phrase can arise in ordinary purchasing discussions without any medical-device or supplier-compliance context, allowing unintended activation despite the skill's specialized description.

Content

No source excerpt is available for this finding.