Back to skill

Security audit

行李清单 Pro

Security checks for vulnerabilities and agentic risk

Overview

The skill is a travel packing assistant, but it asks agents to make unsafe system and network changes that are not necessary for that purpose.

Review this before installing. The packing-list behavior is understandable, but the skill should not require global npm installs, sudo, TLS bypasses, or broad persistent profile reads by default. Prefer a version that pins and verifies dependencies, removes sudo guidance, keeps TLS validation enabled, uses safe structured command invocation, and makes profile storage narrowly opt-in with clear deletion controls.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (5)

T08 · Insecure Dependencies

Error
Location
SKILL.md:43
Finding

Mandatory Installation of a Mutable Global npm Dependency

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Error
Location
SKILL.md:58
Finding

Root-Level Execution of an Unpinned npm Package

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Error
Location
reference/tools.md:28
Finding

Immediate Execution of a Conflicting Unpinned Package Through npx

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:97
Finding

Disabled TLS Certificate Verification for FlyAI Searches

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:99
Finding

Shell Command Injection Through Travel Search Parameters

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Rogue AgentSelf-Modification, Session Persistence
Findings (23)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill instructs users to set NODE_TLS_REJECT_UNAUTHORIZED=0 for FlyAI searches, which disables TLS certificate verification and permits man-in-the-middle interception of network traffic. This is highly unjustified for destination search and exposes queries, results, and potentially credentials or session data to tampering.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger phrases include broad everyday expressions such as '带什么', '打包', and '出行准备', which can cause accidental activation outside the intended context. Misfires can lead to unnecessary data collection, memory access, or tool usage when the user did not intend to invoke this skill.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill reads persistent user travel profile/history at startup before establishing necessity for the current request. This can expose previously shared personal data beyond the immediate task and violates data minimization, especially if the current conversation does not require historical preferences.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill requires a global npm install/upgrade before performing a simple packing-list task, which is unnecessary privilege and capability expansion for the stated purpose. Installing or upgrading code from the network at runtime increases supply-chain risk and can modify the host environment without clear user need, especially when paired with optional sudo guidance.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
97% confidence
Finding

The skill suggests using sudo to perform a global npm install, encouraging root-level execution of network-fetched code for a non-administrative packing-list workflow. If the package, dependency chain, or transport is compromised, this could result in full system compromise.

Content

Scanner excerpt · SKILL.md (reported line 58)May include surrounding context.

md
| 情况 | 处理方式 |
|-----|---------|
| npm 未安装 | 提示用户先安装 Node.js (https://nodejs.org/) |
| 权限不足 | 建议使用 `sudo npm install -g @fly-ai/flyai-cli@latest --registry=https://registry.npmjs.org` 或使用 nvm 管理 Node |
| 网络问题 | 建议用户检查网络或使用国内镜像 `npm config set registry https://registry.npmmirror.com` |

**注意:** 此步骤只在首次使用时执行,后续调用会直接跳过已安装的情况。

Unsafe Defaults

Medium
Category
Tool Misuse
Confidence
99% confidence
Finding

Setting NODE_TLS_REJECT_UNAUTHORIZED=0 establishes an unsafe default that disables certificate validation for outbound connections. In this skill's context, there is no legitimate need to trade transport security for convenience, making interception and response tampering far more dangerous.

Content

Scanner excerpt · SKILL.md (reported line 101)May include surrounding context.

bash
# 搜索目的地相关旅游信息
NODE_TLS_REJECT_UNAUTHORIZED=0 flyai keyword-search --query "[目的地] 旅游 景点"

# 如果用户有明确活动(如潜水、滑雪),搜索相关内容
NODE_TLS_REJECT_UNAUTHORIZED=0 flyai keyword-search --query "[目的地] [活动类型]"

Unsafe Defaults

Medium
Category
Tool Misuse
Confidence
99% confidence
Finding

This repeated use of NODE_TLS_REJECT_UNAUTHORIZED=0 normalizes insecure transport behavior across multiple search paths. Repetition in the instructions makes the unsafe practice more likely to be adopted broadly and increases the attack surface for manipulated search results or stolen metadata.

Content

Scanner excerpt · SKILL.md (reported line 104)May include surrounding context.

NODE_TLS_REJECT_UNAUTHORIZED=0 flyai keyword-search --query "[目的地] 旅游 景点"

如果用户有明确活动(如潜水、滑雪),搜索相关内容

NODE_TLS_REJECT_UNAUTHORIZED=0 flyai keyword-search --query "[目的地] [活动类型]"

text

**分类参考:**

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill expands from packing-list generation into flight, hotel, and attraction search, which broadens data access and tool usage beyond the declared behavior. While not inherently malicious, this scope creep increases the chance of unexpected tool invocation, additional data collection, and user confusion about what the skill is authorized to do.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill instructs saving newly discovered preferences to memory or local files, creating a retention channel for personal data in plain language. Even with a prompt, the design encourages accumulating travel-related profile data without clear minimization, retention limits, or storage protection.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The referenced capability exposes a broad semantic travel search surface spanning hotels, attractions, flights, and trains, while the skill is presented as a packing-list generator. This mismatch can enable unnecessary data access and scope expansion, increasing the chance the skill performs actions or retrieves information beyond what is required for generating a packing list.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The examples and allowed category values are presented in Chinese, and the parameter guidance implies users must supply Chinese keywords and categories to use the skill. Because the document does not offer an alternative language option or explicitly justify that this skill is China-specific, it creates a natural-language locale constraint without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This markdown file presents all user-facing instructions and examples exclusively in Chinese. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is clearly documented and justified, which is not present here.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
97% confidence
Finding

The documentation instructs use of npx @anthropic-ai/flyai-cli@latest, which pulls and executes the latest package version at runtime rather than a vetted, fixed release. This creates a supply-chain risk: if the package is compromised or a breaking/malicious update is published, users or agents following the docs may execute untrusted code immediately.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document defines a cross-platform mechanism to persist a user travel profile across sessions, including personal preferences and history, even though the skill is described as a packing-list generator. This expands data collection and retention beyond what is necessary for the stated function, increasing privacy risk and creating a durable profile that could be read or reused in unrelated contexts.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document operationalizes collection and persistent storage of user travel-profile data in memory and local files but does not include privacy warnings, sensitivity guidance, retention rules, or user rights controls. Without those safeguards, users may not understand that personal data is being stored beyond the current interaction, increasing the chance of inappropriate collection and reuse.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
88% confidence
Finding

Creating and using ~/.flyai as a persistent storage directory enables session persistence of personal data on disk. While directory creation itself is not malicious, in this context it supports retention of user profile information outside the active session, which increases privacy exposure and may be inconsistent with user expectations for a packing-list assistant.

Content

Scanner excerpt · reference/user-profile-storage.md (reported line 73)May include surrounding context.

如果 ~/.flyai/ 目录不存在,需要先创建:

bash
mkdir -p ~/.flyai

文件格式

Ssd 3

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The example file format establishes a durable user travel profile containing personal, familial, preference, and historical destination data. Such a profile is sensitive because it can reveal lifestyle, movement patterns, and household characteristics, and storing it locally may expose it to other local processes or users depending on platform controls.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The unified schema includes broad profile fields such as home city, airport, family composition, travel history, accommodation preferences, and special needs. That set materially exceeds what is required to answer a one-time packing-list request and creates unnecessary sensitivity, especially because family details and travel history can reveal personal patterns.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

These instructions explicitly direct the skill to persist and later reuse personal details in either a memory system or a local file. Cross-session storage of user attributes increases privacy and misuse risk, especially in a skill whose advertised purpose does not require maintaining a lasting personal dossier.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This section instructs the skill to save newly discovered user preferences into Qoder Memory or a local file after prompting for confirmation, enabling long-term retention of personal travel data. For a packing-list skill, that persistence is not necessary to fulfill the immediate task and broadens exposure if the local profile or memory store is later accessed by other components.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The examples and output schema use only Chinese city names, Chinese weekday labels, Chinese duration text, and CNY pricing, which suggests the skill is tailored to a specific language/locale. Because the document does not state that the skill is China-specific or offer any language/locale choice, this can conflict with a policy requiring user opt-in for locale constraints.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The file presents user-facing values and examples in Chinese, such as bed-type values and destination examples, with no indication that other languages/locales are supported. This can violate language-choice policy when a skill effectively forces one language without explicit user opt-in or documented regional scope.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The title and all examples are presented in Chinese, including fixed Chinese field names and example prompts, which can imply a forced language/locale for the skill. There is no indication that the user can choose another language or that the Chinese-only behavior is intentional for a region-specific use case.

Content

No source excerpt is available for this finding.