Back to skill

Security audit

用空位“雷达”替你抢旅游目的地的米其林/热门餐厅

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent restaurant-booking assistant, but it gives the agent under-scoped authority to submit third-party reservations using personal contact details.

Install only if you want an agent to actively monitor booking sites and potentially submit restaurant reservations for you. Give it contact details only after deciding to use it, and require the agent to ask for explicit confirmation before every reservation submission, account login, deposit, card guarantee, cancellation-penalty booking, or payment-related step.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (8)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The '行中模式' example states that the agent has already submitted a reservation through Chope before the user explicitly confirmed that action. In a travel-booking context, this is especially dangerous because it can create unauthorized reservations, trigger financial holds or penalties, and share personal data with a third party without clear consent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

All user-facing instructions, prompts, and output templates are written in Chinese, and the skill does not state that language selection is optional or user-driven. Under the policy, a fixed language requirement without opt-in is a natural-language policy concern unless clearly justified as region-specific.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation criteria are overly broad, including generic travel-planning contexts and many loosely related phrases. This can cause the skill to trigger without clear user intent, leading to unnecessary collection of personal contact data and initiation of reservation-monitoring behavior in contexts where the user did not explicitly request it.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill's stated purpose is restaurant vacancy monitoring, reservation attempts, and notifying the user when a deposit or confirmation is required. However, the privacy constraint explicitly includes '支付信息' as data used in the reservation flow, even though the workflow says payment actions are not automatically completed and should instead be handed off to the user for confirmation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The examples instruct users to provide phone numbers and email addresses and then echo those details back in the agent output, but they do not include any privacy notice, minimization guidance, or explanation of how the data will be stored, shared, or retained. In a reservation-monitoring skill that sends data to third-party booking channels, this increases the risk of unnecessary exposure of personal contact information and normalizes unsafe handling of PII.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The examples describe deposit and credit-card guarantee flows, including no-show penalties, but do not clearly warn the user that proceeding may create a financial obligation via a third-party reservation platform. This can lead users to authorize actions without understanding that a booking may require payment credentials, incur charges, or be subject to cancellation penalties.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
74% confidence
Finding

All headings, prompts, and agent messages are presented in Chinese, with no indication that the user can choose another language. Under the stated policy, forcing a specific language without opt-in can be a natural-language policy issue unless the locale restriction is explicit and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown file presents all user-facing instructions, templates, and guidance only in Chinese. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale restriction is explicitly justified.

Content

No source excerpt is available for this finding.