Security audit
Security checks for vulnerabilities and agentic risk
This skill is a small Douyin-themed demo search helper that returns built-in sample results, with some routing and accuracy caveats but no evidence of hidden access, persistence, exfiltration, or destructive behavior.
Install only if you want a demo/local parser for Douyin-style search requests. Treat returned videos and metrics as sample data unless you add a real API or browser integration, and review any future expansion carefully because real scraping would introduce platform terms, rate limits, network access, and possible credential/session handling.
The skill claims to search and scrape Douyin content but the description itself says the current version only returns simulated results. This mismatch is dangerous because downstream agents or users may treat fabricated output as real external data, enabling misinformation, bad automation decisions, or policy bypass where real retrieval should have been audited.
The README states the skill will automatically trigger for broad, loosely defined requests related to Douyin video search. Overly broad activation boundaries can cause the agent to invoke scraping functionality in unintended contexts, increasing the chance of unauthorized data collection, policy violations, or acting on ambiguous user intent.
The skill advertises scraping/search capabilities and includes examples that imply network access, but it does not declare an explicit tool scope such as allowed tools or permissions. In an agent environment, missing scope boundaries can let the runtime or orchestrator invoke broader network-capable tooling than intended, reducing least-privilege protections and making misuse or unexpected outbound access harder to control.
The trigger phrases are very broad and include generic requests like searching or finding content, which can cause the skill to activate outside a narrowly intended Douyin-specific context. Over-broad activation increases the chance that unrelated user requests get routed into a scraping-oriented skill, causing incorrect behavior, privacy issues, or unintended use of higher-risk tools.
The execution flow says the skill should trigger on common keywords like 'search', 'find', and 'video', which overlaps heavily with everyday language. This creates ambiguous routing and can accidentally invoke a capability associated with scraping or browser automation when a safer, more general skill would have been appropriate.
The example triggers are broad, conversational phrases that overlap with ordinary user speech, which can cause the scraper skill to activate in situations where the user did not clearly intend data scraping or platform-specific search. In a tool that performs external content retrieval, ambiguous activation increases the risk of unintended scraping requests, privacy issues, and misrouting user intent to a data-collection capability.
The top-level docstring presents the file as a natural-language Douyin content search tool, and the CLI description similarly frames it as a functional search utility. In contrast, the code later documents and implements only a simulated result set, creating an active mismatch between documentation and behavior.
The module docstring states the tool in Chinese and gives only Chinese natural-language query examples, which establishes a language-specific interaction model. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is clearly justified, which is not provided here.
The manifest describes a tool for searching and scraping Douyin video data from natural-language requests. However, the implementation explicitly states it is a mock and only matches keywords against an in-memory sample dataset, so the actual behavior does not deliver the claimed search/scraping capability.
The CLI description, help strings, and usage example are all presented only in Chinese, with no alternative language option or explicit notice that the tool is restricted to Chinese-language queries. This creates a language policy issue because users are not offered a choice or informed of a justified locale limitation.
The README emphasizes Chinese natural-language understanding as the supported interaction mode, and all examples are Chinese-only. This can indicate a language/locale constraint without documenting user choice, fallback behavior, or a justified region-specific limitation.
The text says the skill supports Chinese natural language parsing, which implies a language constraint. Because no user opt-in, alternative language support, or justification for the locale restriction is provided, this may conflict with language/locale policy expectations.
All user-facing examples and descriptive text are written in Chinese, and the file does not indicate that other languages are supported or that Chinese is an optional locale choice. This can violate language-choice policy when a skill implicitly constrains interaction to a single language without user opt-in or justification.
No suspicious patterns detected.
