Back to skill

Security audit

全球旅行预订

Security checks for vulnerabilities and agentic risk

Overview

The skill is not malicious, but it overstates live hotel search, room-detail, and booking capabilities that its bundled code does not actually implement.

Review this before installing as a production travel or booking skill. It appears safe from malware-like behavior, but it is currently closer to a stub/demo than a working hotel aggregator; do not rely on it for real prices, availability, room details, or bookings, and avoid granting broad provider credentials until the integrations, consent flow, and booking controls are implemented and documented.

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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill claims real-time aggregation, room queries, and booking across several external platforms, but the file provides only descriptive interface names and no implemented integrations or booking controls. This mismatch can mislead the orchestrating agent or users into assuming verified real-world capabilities exist, increasing the risk of fabricated results, unsafe booking flows, or improper collection/sharing of travel data under false expectations.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding

The skill explicitly requires calling multiple platform APIs and therefore implies network access, but it does not declare any explicit tool scope such as permissions or allowed-tools. This creates an authorization and governance gap: an agent runtime may grant broader network capability than intended, making it harder to constrain which external services can be contacted and to audit cross-platform data flows.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill is designed to send hotel search data across multiple third-party platforms, but it does not warn users that their travel queries may be shared externally. Because travel dates, destinations, and preferences can be privacy-sensitive, the lack of disclosure and consent language increases the chance of silent over-sharing and non-compliant handling of personal data.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The trigger condition says that when a user searches for hotels, the skill should query multiple platforms, which is broad enough to activate on many ordinary travel requests without confirming user intent or consent. In this context, broad invocation is risky because hotel searches often involve dates, destinations, and potentially identity-linked travel preferences that may be sent to several providers unnecessarily.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill tells users to reply with an index to view room details or cross-platform price comparisons, but no corresponding handlers or APIs are implemented in this file. In a hotel booking context, this is dangerous because it can mislead users into believing deeper booking or comparison functionality exists, causing incorrect operational decisions, broken user flows, or unsafe downstream agent behavior if another component improvises unsupported actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The natural-language comments state a strict version requirement and instruct refusal if the version does not match, and the file consistently presents the interface description only in Chinese. Under the stated policy, forcing a specific language/locale without user choice or clear regional justification is a policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

This code file contains natural-language instructions exclusively in Chinese: '版本: 1.0.0 | 最后更新: 2026-03-30' and '如非此版本,请拒绝执行'. In the context of organizational language/locale policy, embedding operational instructions in a single language without offering a user choice can be a locale/language policy concern.

Content

No source excerpt is available for this finding.