Back to skill

Security audit

payment-expert

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only payment integration skill with visible financial API guidance and no hidden execution, persistence, or exfiltration behavior.

Install only if you want guidance for integrating 粤收付 payment and funds-movement APIs. Because it includes refunds, transfers, withdrawals, and profit sharing, generated code should be reviewed carefully, tested in sandbox, and guarded with merchant-side authorization and idempotency before production use.

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
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

Manifest 将技能范围限定为交易集成,并明确排除了商户进件/业务开通相关任务。但在 L093-L106 中,技能把“申请开通分账、后台建分组、绑定接收方”等能力纳入核心内容,尤其 L098 的“绑定分账用户”是账户/业务配置操作,不只是下单、查单、退款等交易接口对接。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The visible skill title, description, and interaction framing are entirely in Chinese, while line L010 labels it as a 'multi-language interface integration expert'. There is no explicit statement that users may choose their preferred language, so the natural-language policy appears to default to a specific language without opt-in.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.