guaikei·小红书博主作品单

• Outcome: Review
• Latest audit: Aug 14, 2026 · 3:58 AM UTC
• Version: 1.0.0

Overview

This skill is a coherent Xiaohongshu public-data retrieval tool, but users should know it sends queries to guaikei.com and saves returned results locally.

Install only if you are comfortable sending Xiaohongshu keywords, note/profile URLs, and your guaikei API token to guaikei.com. Treat the generated logs as retained research data and delete or protect them if they include sensitive campaign, competitor, or personal comment information.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (12)

Description-Behavior Mismatch

Medium

• Confidence: 92% confidence
• Finding: The CLI writes the full fetched comment result set to a local JSON file, which can retain potentially sensitive user-generated content beyond the immediate runtime. In the context of a data-fetching skill, silent persistence increases data exposure risk through leftover files, backups, shared workstations, or later unintended access.

Description-Behavior Mismatch

Medium

• Confidence: 90% confidence
• Finding: The CLI persists the full fetched detail output, including note/comment data and request metadata, to a local file even though the skill is presented as only returning structured data. This creates an undisclosed data-at-rest side effect that can expose scraped content or sensitive business research to other local users, backups, or downstream processes.

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The CLI writes the full search output to a local file after execution, even though the skill is presented as only returning structured Xiaohongshu data. Search terms and results can contain sensitive business intelligence, campaign plans, competitor targets, or user research data; persisting them without clear disclosure or consent increases the risk of unintended local data retention and later exposure to other users, processes, backups, or support workflows.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The description activates the skill for very broad 'data-driven decision' situations, even when the user does not explicitly ask for this skill. That can cause unintended invocation and unnecessary transmission of user queries to a third-party API, increasing privacy and data-minimization risk.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The trigger phrases are everyday, underspecified requests such as '看看XX最近有什么新内容', which can overlap with general research or content-analysis tasks. In an agent environment, such ambiguity can over-route benign user requests into this skill and send terms or links externally without sufficiently informed user intent.

Vague Triggers

Low

• Confidence: 76% confidence
• Finding: Single-note analysis prompts like '这篇笔记怎么样' are ambiguous and can match ordinary summarization or qualitative analysis requests, not necessarily external retrieval requests. The main risk is misrouting and avoidable disclosure of a supplied URL or related context to the third-party service.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill notes third-party relay only in a high-level compliance section and does not clearly state, at the point of normal use, that note URLs, profile URLs, keywords, and query parameters will be transmitted to guaikei.com. This weakens informed consent and may expose user interests, research targets, or monitored accounts to an external provider.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The README states that all task results are automatically saved to a local logs/ directory, but it does not clearly warn users that searched keywords, analyzed links, and returned content may be persisted on disk. In a data-analysis skill handling competitor monitoring, KOL screening, and comment collection, these logs can expose sensitive business intelligence or personal data to other local users, backups, or downstream tooling.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The tool saves comment data to disk without an explicit warning or consent at the moment of persistence, which can surprise users and create unnecessary local storage of potentially sensitive content. Because comments may contain personal data or business-sensitive monitoring outputs, undisclosed persistence increases privacy and data-handling risk.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The program writes detailed task results to a local JSON file without a clear user-facing warning at execution time or in the CLI contract. Because the results may include harvested comments, note details, and operational metadata, silent persistence increases the risk of unintended retention, disclosure, and compliance issues.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The CLI persists the full request and fetched results to a local JSON file, including the supplied profile URL and potentially sensitive scraped data, without any explicit user notice or opt-in. In a data-collection skill, silent persistence increases privacy and data-retention risk because subsequent local users, backups, logs, or other processes may access the stored output unexpectedly.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The tool silently saves search results to a local JSON file without warning the user in its help text or runtime messaging. Hidden persistence is risky because users may assume results are ephemeral, while the saved files can accumulate sensitive queries and harvested content data that may later be accessed by other local users, malware, backups, or automated tooling.

VirusTotal

64/64 vendors flagged this skill as clean.

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