Security audit

酒店采购价格监控

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Chinese hotel-procurement price lookup helper that uses external searches, with some privacy and broad-trigger caveats but no malicious behavior found.

Install only if you are comfortable sending searched product names and procurement interests to external search or marketplace sources. Avoid using confidential SKU codes, unreleased menu plans, supplier strategy, or sensitive internal product names unless the organization approves that disclosure.

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 (10)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The activation phrases are very broad and include common conversational terms like '多少钱' and '价格查询', which can cause the skill to trigger unintentionally during normal user interactions. Because the skill performs broad third-party lookups, accidental activation can lead to unexpected external queries, privacy leakage of user-supplied product terms, and user confusion about why web requests were made.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill defines multiple very broad activation phrases such as '多少钱', '市场价', and '比价', which are common in ordinary conversation and not specific to hotel procurement. This can cause accidental triggering in unrelated contexts, leading the agent to invoke web search/fetch behavior unexpectedly and potentially expose user queries or consume resources when the user did not intend to use this skill.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The script’s comments and all user-facing messages are in Chinese, including usage and status output. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified, which is not present here.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The script sends the user-supplied product name directly to external search services via web_search without any notice, consent, or data-handling disclosure. While the input is only a product term, it may still contain sensitive procurement plans, internal SKU names, or confidential sourcing interests, so silent transmission to third parties creates a real privacy and data-exposure risk.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: This function performs another outbound search using the raw user-provided product term, again without informing the user that their query will be disclosed externally. Repeated transmission across multiple sources increases exposure and can reveal business intent, supplier interest, or confidential purchasing activity.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The e-commerce lookup sends the user's search terms to external services covering multiple marketplaces without explicit disclosure. In a hotel procurement context, searched product names can reflect planned inventory, menu changes, or vendor strategy, so leaking them to third parties may expose commercially sensitive information.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The cost-analysis query forwards the product term together with cost/margin-oriented keywords to external search services, which can reveal even more sensitive internal business analysis than a basic price check. This makes the issue more dangerous in context because it may disclose strategic sourcing, margin investigation, or BOM-related research to third parties.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 85% confidence
Finding: The README presents the skill entirely in Chinese and all activation examples are Chinese-only, with no indication that users may choose another language or that the locale restriction is intentional and justified. This can be a natural-language policy issue when a skill implicitly enforces a specific language without opt-in.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The README advertises multi-source concurrent queries across numerous third-party platforms but does not warn users that their queries may be sent externally. This reduces transparency and informed consent, especially in a procurement setting where searched items, pricing interests, or sourcing patterns may be commercially sensitive.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 89% confidence
Finding: All user-facing strings, examples, and usage text in the script are in Chinese, which imposes a specific language on users. There is no indication that the skill is region-specific or that users may opt into another language.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.