Back to skill

Security audit

外卖运营诊断台

Security checks for vulnerabilities and agentic risk

Overview

This is a simple food-delivery operations advice skill with no code execution, credential access, network behavior, or persistence.

Reasonable to install for Chinese-language food-delivery operations help. Avoid entering credentials, private customer details, or unnecessary personal data, and verify any platform-rule or pricing decisions against current official platform guidance.

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
  • 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
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger examples are short, common phrases such as '外卖运营', '曝光低', and '下单少', which are broad enough to match ordinary conversation outside a deliberate skill invocation. This can cause unintended activation and inject the skill's behavior into unrelated chats, increasing the chance of context confusion or accidental processing of user business data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The natural-language examples are open-ended and resemble normal user requests rather than clearly bounded invocation syntax. Because they do not distinguish between casual discussion and intentional tool use, the orchestrator may activate the skill unexpectedly, leading to misrouting or unauthorized use of supplied context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

All user-facing descriptions, parameters, trigger examples, and safety notes are presented only in Chinese, which effectively enforces a single language experience. The file does not state that the skill is region-specific or give users an option to request another language.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.