Security audit

guaikei-douyin-all-public-data

Security checks across malware telemetry and agentic risk

Overview

The skill does what it says, but it can broadly collect Douyin public data, send user queries to an external API, and automatically save large datasets locally with limited safeguards.

Review this skill before installing if your Douyin research terms, target accounts, or comments could be sensitive. Use a dedicated API token, avoid running it on shared machines unless you can protect the logs directory, delete exported JSON files when no longer needed, and confirm that any bulk collection complies with applicable law, platform terms, and your organization’s data-handling rules.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (9)

Description-Behavior Mismatch

Medium

Confidence: 94% confidence
Finding: The skill states it does not write to the filesystem except logs, but later says outputs are automatically saved under `logs/`. That inconsistency can mislead operators about data persistence and may cause sensitive scraped data, query terms, or derived results to be retained on disk unexpectedly, increasing exposure to local disclosure or overcollection.

Context-Inappropriate Capability

Low

Confidence: 92% confidence
Finding: The authentication failure path includes an unrelated out-of-band contact channel and apparent upsell message ('联系微信...获取解决方案') inside a security-relevant error. This is risky because it can socially engineer operators into contacting an external party for credential issues, bypass normal support processes, and leak operational details; in a data-collection skill context, that increases trust and supply-chain concerns.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The generalized trigger allows the skill to activate for broad research requests even when the user did not explicitly ask for Douyin data. Over-broad invocation can cause unintended external data access, collection of platform-specific data without clear consent, and execution in contexts where a narrower or different tool would have been more appropriate.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The README states that scraped data is automatically exported to local log files, including comments and account-related public data, but it does not clearly warn users that this creates a persistent local copy of potentially sensitive or regulated content. In the context of social-media scraping and comment analysis, silent disk retention increases the risk of unintended disclosure, over-retention, and secondary misuse of personal data.

Missing User Warnings

Low

Confidence: 82% confidence
Finding: The quick-start requires users to place `GUAIKEI_API_TOKEN` in an environment variable but gives only minimal advice about protecting the token. While environment variables are common, failing to clearly describe the token as a secret can lead to accidental exposure through shell history, screenshots, shared terminals, CI logs, or inherited process environments.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The documentation explicitly supports bulk collection of Douyin creator posts and comments, but provides no warning about privacy expectations, data protection, rate limits, or platform terms. In a data-scraping skill, this omission can normalize collection and downstream processing of user-generated content without safeguards, increasing the risk of misuse, overcollection, and non-compliant handling of public personal data.

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The CLI writes the full comment retrieval results to a local JSON file automatically after execution, without explicit user confirmation or a prominent warning. Because comment data may contain personal information, usernames, or other sensitive content derived from public scraping, this creates a data-handling risk: users may unknowingly leave recoverable artifacts on disk, exposing collected data to other local users, backups, or unintended sharing.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The CLI persists the full search output to a local file named with the user keyword and parameters, but it does not clearly warn the user that their query and returned public-data results will be stored on disk. This creates a privacy and operational risk because sensitive or regulated research terms can be retained locally, exposed to other users on the host, or later collected by backup/logging systems without the operator realizing it.

Missing User Warnings

Medium

Confidence: 82% confidence
Finding: The tool sends user-supplied search terms and filters to external API functions without an explicit disclosure that the data will leave the local environment. In a data-collection skill, this is expected behavior, but the absence of clear notice can still cause unintentional disclosure of sensitive business research, campaign terms, or investigator queries to the remote service.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.