Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md - `-32601` (Method not found): 检查 `SKILL.md` 中定义的工具名是否在 MCP 侧已注册。
Security audit
Luogang shopping assistant
Security checks for vulnerabilities and agentic risk
Overview
This skill is a straightforward Luogang shopping lookup helper that sends product searches to a disclosed external MCP service, with no local persistence or hidden privileged behavior found.
Install only if you are comfortable with shopping searches and product-detail requests being sent to the Luogang MCP service. For generic shopping questions, confirm you want Luogang results before the assistant uses this skill.
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
- 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 (6)
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The skill description is broadly scoped to activate on general shopping intent and multiple brand aliases, which can cause the assistant to invoke this skill for ordinary conversation that merely mentions buying or product lookup. Overbroad activation increases the chance that unrelated user queries are routed to this skill and then transmitted to the external MCP service without clear necessity or informed user expectation.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 92% confidence
- Finding
The skill explicitly instructs the assistant to call a remote MCP service for real-time product data, but it does not include a user-facing warning that user queries and product interests may be sent to an external service. This is a privacy issue because shopping intent, preferences, and possibly identifiers can be disclosed to a third party without transparent notice or consent.
- Content
External Transmission
Medium
- Category
- Data Exfiltration
- Confidence
- 96% confidence
- Finding
The documented curl invocation sends user-supplied arguments to an external HTTP endpoint, which is a real external-transmission behavior rather than a purely local skill. While this appears functionally necessary for the shopping integration, it still creates security and privacy risk because user queries, product interests, and potentially store/channel identifiers are transmitted off-platform to a third-party service.
- Content
调用方式
bash curl -s -X POST "$LUOGANG_MCP_HTTP_URL" \ -H "Content-Type: application/json" \ -H "Accept: application/json, text/event-stream" \ -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"<工具名>","arguments":{<参数>}},"id":1}'
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The trigger examples include very generic phrases like '找外套', '买鞋子', and '有没有外套', which overlap heavily with normal conversation and could cause this skill to activate far outside its intended Luogang-specific context. In practice, this can silently expand external data sharing and tool usage whenever a user asks about products, even if they did not intend to use this third-party shopping service.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The detail-view triggers such as '查看详情', '这个多少钱', and '还有货吗' are ambiguous because they could refer to any item currently being discussed, not necessarily a Luogang product. This creates a risk of unintended skill activation and unnecessary external requests, especially in multi-skill or general-assistant contexts where such phrases are common follow-ups.
- Content
Static analysis
No suspicious patterns detected.