Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed Chinese daily-news report generator that searches the web, formats a report, and saves a local archive copy.
Install this if you want a Chinese daily news report that performs multiple web searches and keeps local report archives. Review or disable the cron example and delete memory/daily_news_*.md files if you do not want persistent historical reports.
The skill’s natural-language instructions and reporting template are entirely specified in Chinese, and the cron payload also invokes the skill in Chinese. There is no indication that users may choose another language or that the Chinese-only behavior is a documented, justified locale restriction.
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
"schedule": { "kind": "cron", "expr": "30 8 * * 1-5", "tz": "Asia/Shanghai" },
"payload": {
"kind": "agentTurn",
"message": "按 skills/daily-news/SKILL.md 执行每日新闻日报 v2.3"
},
"timeoutSeconds": 900
}
Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.
"schedule": { "kind": "cron", "expr": "30 8 * * 1-5", "tz": "Asia/Shanghai" },
"payload": {
"kind": "agentTurn",
"message": "按 skills/daily-news/SKILL.md 执行每日新闻日报 v2.3"
},
"timeoutSeconds": 900
}
The natural-language description is entirely in Chinese and provides no indication that users can choose another language or locale. Under the policy, forcing a specific language without user opt-in is a reportable locale-policy issue unless the constraint is clearly documented and justified.
The skill explicitly instructs saving the full generated report to memory/daily_news_YYYY-MM-DD.md without any notice, consent flow, or retention limit. Persistent storage of outputs can unintentionally retain sensitive user context, proprietary prompts, or browsing-derived content beyond the immediate session, creating avoidable privacy and data-governance risk.
No suspicious patterns detected.