guaikei-douyin-get-hot-ranking

• Outcome: Review
• Latest audit: Aug 18, 2026 · 2:42 AM UTC
• Version: 1.0.0

Overview

This Douyin data skill is mostly coherent, but it needs review because it can bulk collect and locally retain public user/comment data and may expose media download URLs despite saying downloads are unsupported.

Review this before installing if you handle personal data or compliance-sensitive research. Use it only for public Douyin data you are allowed to collect, expect requests and tokens to go to guaikei.com, and treat generated logs as retained datasets that may contain comment text, user IDs, nicknames, profile links, and IP-region labels. Also be aware that outputs may include direct media URL fields even though the skill says it does not provide download capability.

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 (10)

Intent-Code Divergence

Medium

• Confidence: 96% confidence
• Finding: The schema explicitly documents `video_url` as containing both playback and download addresses, which conflicts with the skill metadata claim that it does not provide download capability. This discrepancy can mislead downstream integrators or agent logic into exposing or relying on a download feature, creating a policy bypass and increasing the chance of unauthorized content acquisition.

Description-Behavior Mismatch

Medium

• Confidence: 86% confidence
• Finding: The schema explicitly documents `play_addr` as both a playback and video download address, which conflicts with the skill’s stated restriction that it does not provide download capability. In a scraping/search skill for Douyin content, exposing a downloadable media URL can enable unauthorized content retrieval and policy bypass even if the manifest claims otherwise.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The description uses broad trigger language like '当用户想搜、想扒、想翻、想追抖音上的东西时' and even casual phrasing examples, which can cause the orchestrator to invoke this skill for loosely related conversation. Over-broad routing increases the chance of unnecessary third-party data disclosure, unintended API usage, and execution in contexts where the user did not clearly request Douyin data retrieval.

Vague Triggers

Low

• Confidence: 91% confidence
• Finding: The natural-language trigger examples include broad colloquial requests such as '抖音今天有什么热点' and '帮我做抖音竞品分析' with limited guardrails, which can bias a planner toward auto-invocation without first confirming scope or inputs. In this skill's context, that can lead to accidental calls to an external API and collection of data beyond what the user specifically intended.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The skill defines broad natural-language triggers such as colloquial requests about 'what's happening on Douyin' that can match ordinary conversation without requiring clear user intent to invoke data-collection actions. This increases the chance of unintended activation, causing the agent to query a third-party service, use the configured API token, and retrieve platform data when the user did not explicitly request tool use.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The intent recognition priority treats generic phrases like 'hot search', 'hotspot', 'ranking', or 'what is popular today' as sufficient to run the hot-ranking command. Because these phrases are common in everyday chat and are not constrained to Douyin or even to a request for tool execution, the agent may trigger external data access unexpectedly and leak user intent into third-party queries.

Missing User Warnings

Medium

• Confidence: 87% confidence
• Finding: The changelog explicitly advertises bulk scraping of competitor accounts and collection of video comments, but provides no warning about privacy, platform terms, consent, retention, or downstream handling of collected data. In the context of a user-facing scraping skill, this omission can normalize high-volume data collection and enable misuse against third parties, increasing legal, privacy, and compliance risk.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: This documentation explicitly enables bulk retrieval of Douyin creator posts and video comments, which are user-generated content and may contain personal data, without any notice about privacy, lawful basis, consent, rate limits, or downstream handling. In a scraping/analytics skill, omission of those safeguards increases the risk of misuse for profiling, surveillance, harassment, or non-compliant data collection.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The CLI writes fetched comment data to a local JSON file automatically, without any explicit notice, consent, or opt-in at the point of execution. Because comment contents may include personal data, usernames, or sensitive text, this creates an unintended local data retention risk and can expose scraped data to other local users, backups, or downstream processes.

Natural-Language Policy Violations

Low

• Confidence: 78% confidence
• Finding: The authentication error message embeds a direct personal contact identifier (a WeChat/phone number) and instructs operators to contact it when auth fails. This can expose unnecessary sensitive contact information, create a social-engineering support channel outside normal controls, and leak operational details to untrusted users if surfaced in responses or logs. In a networked scraping skill that depends on an API token, such fallback contact instructions are more suspicious than a normal localized error because they encourage out-of-band recovery.

VirusTotal

VirusTotal findings are pending for this skill version.

Static analysis

No suspicious patterns detected.