Security audit

Guaikei Douyin Competitor Watch

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its Douyin analytics purpose, but it uses an external service, saves collected data locally by default, and its runtime token-error messages contradict its stated no-promotion rule.

Install only if you are comfortable with Douyin queries, URLs, and returned public data being processed by guaikei.com and saved in local log files by default. Treat the token like a credential, and review the runtime contact/website behavior before use because it conflicts with the skill's own safety claims.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (8)

Description-Behavior Mismatch

Medium

Confidence: 88% confidence
Finding: The changelog advertises capabilities beyond the stated skill scope, including general Douyin search, hot-search queries, and bulk retrieval of an author's published works. This mismatch can enable scope creep at runtime, making it easier for an agent or operator to invoke broader collection behavior than users or reviewers expect, especially in a competitor-monitoring context involving large-scale scraping and monitoring.

Description-Behavior Mismatch

Medium

Confidence: 92% confidence
Finding: The CLI writes the full request metadata and returned search results to a local file without any visible consent flow, retention controls, or indication in the skill description. This creates an unintended data persistence channel that can expose user queries and collected platform data to other local users, backups, or later misuse.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The activation guidance says to use the skill even when the user does not explicitly mention Douyin, and the trigger phrases include broad, common requests like checking what people said or whether a creator is popular. This can cause overbroad invocation, unintentionally sending user queries to an external data service and producing actions outside the user's clearly expressed platform scope.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The example trigger phrases are generic everyday language and are not tightly bound to Douyin-specific intent. In an agent setting, this increases the chance of accidental activation and unnecessary disclosure of user prompts or search targets to the third-party backend described by the skill.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The CLI persists fetched comment data to a local JSON file automatically after successful execution, without any consent prompt, opt-in flag, or visible disclosure in this code path. Comment datasets can contain personal data, user-generated content, and potentially sensitive business intelligence; silently writing them to disk increases the risk of unintended retention, later exfiltration, or access by other local users/processes.

Missing User Warnings

Medium

Confidence: 86% confidence
Finding: The CLI persists fetched results to a local JSON file automatically, which can create unintended data retention and disclosure risks on shared machines, CI runners, or multi-user environments. Because the output may contain scraped account/post metadata tied to user requests, silent storage increases the chance of later unauthorized access or mishandling.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: This code sends a user-supplied Douyin URL together with an API-backed authenticated request to an external service, but the user-facing behavior does not clearly disclose that their input is transmitted off-box. In an analytics skill context, hidden third-party transmission can expose user queries and targets to external processing, creating privacy and trust risks even if the transmission is functionally required.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: Search queries and results are persisted locally without an explicit user-facing warning, even though they may reveal interests, research targets, or monitored accounts. Silent local retention increases privacy risk and can leak data through shared systems, endpoint collection tools, or accidental file disclosure.

Static analysis

No suspicious patterns detected.