Security audit

guaikei-rednote-insight

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Xiaohongshu public-data research wrapper, with the main cautions that it sends queries to guaikei.com and saves results locally.

Install only if you are comfortable sending Xiaohongshu keywords, URLs, and your GUAIKEI API token to guaikei.com. Use a token obtained through a channel you trust, avoid submitting sensitive URLs or business research from shared environments, and periodically delete or protect the local logs directory because results are saved automatically.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Intent-Code Divergence

Low

Confidence: 95% confidence
Finding: The token helper does more than validate a token: when validation fails, it directs the user to contact a specific WeChat number to obtain a 'private token'. In an agent skill context, this is suspicious social-engineering behavior because it funnels users toward an off-platform secret distribution channel, potentially enabling credential harvesting, unauthorized access brokering, or monetized backdoor access.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The skill description explicitly says it should be used even when the user only asks broad questions like '最近什么火' or '帮我找热门内容' as long as the context is social-media mining. This creates an over-broad trigger surface that can cause the agent to invoke this Xiaohongshu-specific skill for ambiguous requests, potentially sending user queries to a third-party API without clear platform intent or informed consent.

Missing User Warnings

Low

Confidence: 88% confidence
Finding: The README states that all task results are automatically saved to the local `logs/` directory, but it does not present this as a prominent user warning with privacy and retention implications. Because the tool processes URLs, keywords, comments, and competitive monitoring data, automatic persistence can unexpectedly retain sensitive business research or personal data on disk.

Missing User Warnings

Medium

Confidence: 83% confidence
Finding: The tool persists the full request URL and API results to a local JSON file without explicit user consent or a visible opt-in. Because the profile URL may include query parameters such as xsec_token and the results may contain scraped account data, this creates avoidable local disclosure risk on shared systems or in environments where working directories are later synced, backed up, or committed.

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