Security audit
Security checks for vulnerabilities and agentic risk
This skill is a Chinese-language daily briefing tool that openly sends generated reports to Feishu when configured, with no evidence of hidden persistence, credential theft, destructive behavior, or unrelated data access.
Before installing, confirm that you want briefings sent to the configured Feishu chat and that the chat is appropriate for any topics you request. Review or remove feishu_chat_id if you only want local report generation, and be careful with broad trigger phrases in shared or automated contexts.
The description emphasizes report generation but does not clearly warn users that content may be automatically pushed to Feishu. This omission matters because users may invoke or configure the skill without realizing it performs outbound messaging, which can leak sensitive interests, create unwanted notifications, or send content into shared workspaces.
The trigger phrases include very broad, natural-language expressions such as '今天早报' and '生成早报', which are likely to appear in ordinary conversation and can cause accidental invocation. In this skill, unintended activation is more dangerous because execution can generate content and automatically push messages to Feishu, creating privacy, spam, and workflow-disruption risk.
The natural-language metadata is entirely in Chinese, and the file does not document that the skill is China-specific or that users can opt into this locale. Under the policy, forcing a specific language without user choice or justified regional scope is a natural-language policy violation.
The file's comments, topic matching, and user-visible success message are all Chinese-specific, and topic detection only recognizes Chinese keywords such as 财经, 科技, and 股市. This imposes a language/locale constraint in the skill behavior without offering the user a language choice or documenting that the skill is intentionally limited to a Chinese-language context.
The handler automatically sends generated content to a Feishu chat whenever feishu_chat_id is present in local configuration, without any runtime user consent, preview, or warning. This creates a real data exfiltration/privacy risk because user-provided text influences the generated report and the skill can transmit that content to an external messaging service silently.
