Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed cron scheduler, but it can create persistent background tasks and automatically send potentially sensitive summaries to Telegram or WhatsApp without enough safety framing.
Review this skill carefully before installing. Use it only for explicit scheduling requests, confirm every recurring or isolated-session job, avoid sending email/calendar/project summaries to external channels unless recipients and content are reviewed, and check/remove created cron jobs when no longer needed.
The description states '支持中文交互' and the document is presented as centered on Chinese interaction, but it does not indicate that users may choose another language. This can be a locale or language policy issue when a skill implicitly enforces a specific interaction language without opt-in.
The skill documents sending task results to external channels such as Telegram and WhatsApp, but it does not clearly warn that scheduled content may leave the local agent environment and be transmitted to third-party services. This creates a real risk of accidental data exfiltration, especially because scheduled jobs may summarize emails, calendars, or project progress and then deliver that information automatically to external recipients.
The surrounding documentation describes a cron scheduler configured with parameters such as --at, --cron, --session, --message, --channel, and --to. In contrast, lines L090-L097 describe unrelated generic workflow inputs, which actively conflicts with the actual stated task interface and can mislead users about what the skill accepts.
Lines L098-L144 present a concrete JSON response schema with fields like success, data, and execution_log, implying the skill itself returns this structure. However, the rest of the file documents manual CLI usage for scheduling tasks rather than an implementation that produces that JSON output, so the documentation misstates what the skill does.
The trigger condition is broad enough to auto-activate this skill for many vague automation requests, increasing the chance that an agent will invoke scheduling or background execution when the user did not explicitly ask for it. In a skill that supports exec and persistent scheduled tasks, overbroad routing raises the risk of unintended task creation, unintended persistence, and downstream actions occurring outside the immediate conversation context.
No suspicious patterns detected.
