Security audit
Security checks for vulnerabilities and agentic risk
This skill openly describes setting up an automated daily news digest with scheduled jobs, web searches, a local cache file, and a Tavily API key.
Before installing, confirm that you want recurring scheduled jobs, know where the weekly news file and cron entries will be created, and store the Tavily API key only in an appropriate environment or secrets location. Avoid configuring sensitive or private search topics because queries are sent to an external search provider.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
## Setup
1. Sign up at https://tavily.com
2. Get API key from dashboard
3. Set environment variable: `TAVILY_API_KEY=tvly-dev-xxxxx`
## Usage
The trigger list is broad and includes common phrases like "daily news," "news digest," and "automated newsletter," which can overlap with ordinary user requests rather than clearly signaling use of this specific automation skill. This can cause the agent to invoke the skill unexpectedly, leading to unintended web access, scheduled task setup, or file-writing behavior in contexts where the user only wanted general advice or a simple summary.
The source list explicitly includes multiple Chinese-language outlets such as InfoQ.cn, 36氪, 机器之心, 量子位, 虎嗅, and 雷科技 as standard inputs, but the document does not state that Chinese-language sourcing is optional, user-selected, or required for a region-specific purpose. This can impose a locale/language constraint implicitly, which matches the policy category for forced language or locale without user opt-in.
The guide tells users to obtain a Tavily API key and have the AI configure it as an environment variable, but it omits guidance on secure secret handling and the privacy implications of sending search queries to a third-party service. This creates risk of credential leakage, overbroad key exposure in the runtime environment, and inadvertent disclosure of sensitive topics or internal interests through external queries.
The guide explicitly instructs the AI to set environment variables, create cron jobs, write files, and send automated digests, but it does not mention obtaining explicit user confirmation, disclosing persistence, or warning that the system will be modified on an ongoing basis. In an agent setting, this can lead to silent persistence, unintended background activity, and surprise use of local resources or credentials beyond the user’s immediate request.
The customization instructions rely on Chinese phrases as the example values to edit, which can impose a language expectation on users who were otherwise addressed in English. Because no language option or justification is provided, this can be interpreted as a locale/language policy issue.
No suspicious patterns detected.