Security audit

guaikei·小红书笔记详情

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Xiaohongshu public-data analysis tool, with manageable privacy considerations around third-party API use and saved local result logs.

Install only if you are comfortable sending Xiaohongshu keywords or URLs and your GUAIKEI_API_TOKEN to guaikei.com. Review or delete the generated logs directory when results may contain sensitive research topics, account links, comments, or competitive monitoring data.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Description-Behavior Mismatch

Medium

Confidence: 88% confidence
Finding: The CLI writes fetched results to a local JSON file automatically after successful execution. Even if the source data is public, persisted output can contain search history, targeted profile URLs, timestamps, and aggregated results that create a local privacy and data-retention risk, especially on shared systems or when logs are collected elsewhere.

Description-Behavior Mismatch

Medium

Confidence: 84% confidence
Finding: The CLI persists full search output to a local JSON file without clearly disclosing that behavior in the skill description or CLI flow. Search results and user-provided keywords can contain sensitive research topics or operational context, and writing them to disk creates unintended data retention and a secondary exposure surface on shared systems.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: The skill explicitly supports retrieving commenter information and creator post histories from a third-party API, but the main workflow sections do not prominently warn the user about privacy implications or data externalization at the point of use. This can lead to collection and transfer of personal data without clear informed consent, especially when analyzing individuals rather than aggregate trends.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: Automatically saving all task results to disk can create unintended local data exposure, especially when retrieved content includes comments, account data, or commercially sensitive monitoring results. Even if the source data is public, persistent storage without a clear warning, retention policy, or opt-out increases the risk of accidental disclosure on shared systems or through backups.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The program silently writes fetched results to a local JSON file without an explicit user-facing warning or consent mechanism. This can expose collected data and usage metadata to other local users, backup systems, or log harvesters, which is more risky than console-only output because it creates durable artifacts.

Missing User Warnings

Low

Confidence: 80% confidence
Finding: The program writes search results to a local JSON file but does not give an explicit warning in its CLI behavior that user queries and returned content will be stored on disk. On multi-user machines or monitored environments, that can leak a user's interests, investigative topics, or collected content beyond the immediate session.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical

Code: suspicious.exposed_secret_literal
Location: SKILL.md:16