Security audit

guaikei·小红书搜索详情评论

Security checks across malware telemetry and agentic risk

Overview

This skill is a purpose-aligned Xiaohongshu public-data lookup tool, but users should understand it sends queries to guaikei.com and saves returned results locally.

Install only if you are comfortable sending Xiaohongshu keywords, note URLs, profile URLs, limits, and your GUAIKEI_API_TOKEN to guaikei.com. Review or delete the generated logs directory if saved public social-media data or comments should not persist locally.

SkillSpector

By NVIDIA

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
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 83% confidence
Finding: The skill requires access to an API token via environment variables, but no explicit permission declaration or user-facing permission framing is present. This creates a transparency and governance gap: agents may invoke a networked capability using sensitive credentials without clear policy scoping, increasing the risk of unintended external data access or token misuse.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger conditions are overly broad, including cases where users do not explicitly mention Xiaohongshu but reference related terms or links. This can cause the skill to activate on ambiguous requests and send user-provided content or URLs to the third-party API unexpectedly, creating privacy, consent, and data-minimization risks.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: Although the document notes dependency on a third-party API, the main workflow and usage sections do not prominently warn that note links, profile URLs, keywords, and other query content will be sent to guaikei.com. Users or orchestrating agents may therefore disclose sensitive research targets or private business intent to an external processor without informed awareness.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The CLI persists scraped Xiaohongshu results to a local JSON file automatically after printing them, but this file does not clearly warn the user that content and metadata from a public profile will be stored on disk. Local persistence increases the risk of unintended retention, later disclosure to other local users/processes, and accidental inclusion in backups, sync tools, or source control. In this skill context, the data is scraped social-media content rather than highly privileged secrets, which lowers severity somewhat, but comments/profile data can still contain personal information and create privacy/compliance exposure.

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:17