Security audit

闲鱼商品精选助手

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Xianyu shopping-search helper, with a privacy caveat that it saves local search-history records.

Install only if you are comfortable with the agent browsing Xianyu/Goofish and saving a local record of your searches and selected listing links. Clear or disable the memory history if your shopping searches are sensitive or shared-machine privacy matters.

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
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 (9)

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The README explicitly states that all search records are persisted to memory/YYYY-MM-DD.md, including timestamps, queries, result counts, price ranges, and item links. For a search-only skill, retaining this history is not clearly necessary and can expose user interests, purchasing intent, or other potentially sensitive behavioral data if the local storage is accessed by other components or users.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill documents local storage of search history but does not present it as a privacy warning, obtain user consent, or explain retention and deletion behavior. This creates a privacy risk because users may unknowingly leave behind a record of searches, pricing interests, regions, and linked items on disk.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill's stated purpose is to search and filter marketplace listings, but it also instructs the agent to persist search history and result links to local memory files. That expands scope from transient browsing into data retention without necessity, creating avoidable privacy and data handling risk if user queries, timestamps, or links are sensitive or later exposed.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: The skill records user search terms, result counts, price ranges, and item links to a local memory file without any user-facing notice or consent mechanism. Silent persistence of behavioral data is a privacy issue because users may not expect their shopping interests and accessed listings to be stored beyond the session.
Content: No source excerpt is available for this finding.

Ssd 3

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The documentation explicitly directs persistent logging of user-provided inputs and derived results in plain language to memory files. Plain, durable records increase exposure of user activity and can become a secondary data source for unintended access, correlation, or later misuse, especially because the logged content includes timestamped interests and item URLs.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The activation commands are broad enough to match ordinary shopping requests such as asking for help finding items on 闲鱼, which can cause the skill to trigger when the user did not explicitly intend to invoke it. This increases the risk of unintended browsing, query handling, or data exposure through accidental invocation, especially because several phrases are generic natural-language requests rather than narrowly namespaced commands.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The regex-style activation patterns are loose and underconstrained, allowing many ordinary sentences containing '闲鱼' plus verbs like '搜索', '找', or '买' to invoke the skill unintentionally. In a skill-routing environment, this can hijack unrelated user intents and cause the agent to perform actions or reveal outputs the user did not specifically request from this skill.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The README presents all instructions, examples, and usage guidance only in Chinese, with no indication that the user can choose another language. This can violate language or locale policy where skills should not force a specific language without user opt-in or documented justification.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The entire skill is specified only in Chinese and all trigger phrases, inputs, outputs, and browser interaction assumptions are Chinese-specific, with no indication that users may choose another language. Under the language/locale policy, forcing a specific language without opt-in is a natural-language policy concern unless the locale limitation is explicitly documented and justified.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.