Skip to content

Expecho/Agent42 / src/Agent42.Agents/Skills/scheduledprompt

scheduledprompt

Verified

Schedule and cancel prompts for the current conversation with the schedule_prompt, cancel_scheduled_prompt, and cancel_all_scheduled_prompts tools. Use whenever the user asks to be reminded of something, wants a recurring nudge, wants the agent to do something later, or wants to stop, cancel, remove, or clear a reminder — for example "remind me to stand up every 30 minutes", "check in on me every morning", "cancel my stand-up reminder", or "clear all my reminders".

Reminders

You can schedule prompts for the current conversation with the schedule_prompt tool. A prompt is a scheduled instruction: when it fires, its instruction is sent to you as a user message in this conversation, and you act on it — your response is saved in the conversation history for the user to see.

When to use it

Call schedule_prompt whenever the user asks to be reminded about something, wants a recurring nudge, or wants you to do something later — for example "remind me to stand up every 30 minutes", "check in on me every morning", or "summarize this conversation in an hour".

How reminders behave

  • They are backed by Orleans reminders, so they are durable: a reminder survives the agent being deactivated and the silo restarting, and it belongs to this conversation.
  • When a reminder fires, its prompt arrives as a user message while the user may be away; answer it as a normal turn, addressed to the user, so it reads well when they come back.
  • A reminder first fires after an initial delay and then repeats on a fixed interval; Orleans reminders always recur, there is no true one-shot.
  • The smallest granularity Orleans supports is one minute, so both the initial delay and the repeat interval must be at least 1 minute. If the user asks for something faster, explain that one minute is the minimum.

Arguments to provide

  • name — a short, stable identifier such as stand-up or daily-check-in. Reusing an existing name updates that reminder instead of creating a duplicate.
  • prompt — the instruction you will receive when the reminder fires. Write it as a self-contained instruction to your future self, e.g. "Remind the user to stand up and stretch." or "Write a short summary of the conversation so far."
  • dueInMinutes — how long from now the reminder should first fire (must be ≥ 1).
  • repeatEveryMinutes — how often it repeats after the first firing (must be ≥ 1). For a reminder the user wants only once, use a large interval and tell them it will keep repeating until it is cancelled.

After scheduling

Confirm back to the user exactly what you scheduled — when it will first fire, how often it repeats, and what will happen when it does. Never claim a reminder was set unless the tool call actually succeeded; if the tool returns an error message, relay it and ask the user how to adjust.

Cancelling

To stop a reminder, use the cancel tools:

  • cancel_scheduled_prompt — cancels one reminder. Pass its name, the same stable identifier used when it was scheduled (e.g. stand-up). Use this when the user wants to stop a specific reminder. If the name doesn't match a scheduled reminder, the tool says so — relay that rather than claiming success.
  • cancel_all_scheduled_prompts — cancels every reminder in this conversation at once. Use it only when the user clearly wants to clear all of them, never to cancel a single one.

Both permanently unregister the durable reminder(s) so nothing fires again. After cancelling, confirm to the user exactly what was cancelled, and only claim a reminder was cancelled if the tool call actually succeeded.

reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 470kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall