Casual Cron

**Quick judgment** Casual Cron is a focused and well-scoped skill for translating natural-language scheduling requests into OpenClaw cron jobs, with strong safeguards for how scheduled jobs run. It is a good fit whenever users want reminders or timed messages (especially via Telegram and `/at`/`/every` commands) and you need to express them as `openclaw cron add` CLI calls. **Strengths & advantages** - Handles both one-shot and repeating schedules, including interval durations and cron expressions. - Enforces safe execution within cron jobs: no troubleshooting, no gateway restarts, no time checks, and output of **only** the message payload when the cron fires. - Uses isolated sessions, delete-after-run for one-shots, and explicit `--deliver --channel --to` flags, which limits side effects and clutter. - Includes a confirmation step (time, job name, job id) before creating jobs, reducing the risk of wrong schedules. - Supports multiple channels (telegram, whatsapp, slack, discord, signal) and offers DST-aware behavior via `America/New_York`. **Risks & limitations** - Natural-language time parsing can be ambiguous; misinterpretation of times, dates, or recurring patterns is the main reliability risk. - The default timezone (`America/New_York`) may be surprising or incorrect for users in other regions if not clearly surfaced and confirmed. - Users must supply correct channel/destination identifiers; mistakes there can lead to messages going to the wrong place or failing silently. - In cron context, the strict requirement to output only the payload (no explanations) means you must design the payload carefully up front, as you cannot add runtime clarification. **Recommended scenarios** Use Casual Cron when you explicitly detect scheduling or reminder intent—especially with `/at` and `/every` commands or phrases like “remind me in 20 minutes”, “daily at 8am”, “every 2 hours”, or “send me a Telegram message at 3pm”—and you need robust, low-risk cron jobs that send precise chat messages at the specified times.