Agents that ask for approval while you're away: what actually happens to the request?
retained text
Over the past week I've seen at least three people build their own fix for the same thing: a coding agent sitting idle on a permission prompt while the human is in another window. One turned sessions into villagers that walk up and ask "May I?", one put the prompts in the Mac notch, one forwards them to a Telegram bot. Those all solve the "I'm at my laptop but not looking" case. The case I can't find good answers for is when nobody is at a laptop: \- an event-triggered job fires at 11pm and wants sign-off on a refund or a budget change \- the person who owns it is asleep, on a flight, or has their phone on do not disturb \- the agent is correctly waiting, so nothing breaks. it just quietly doesn't happen Most setups I've read about send the ask to Slack or email, then default to "wait" or "expire" when nobody answers. That's safe, but it means the approval gate is only as fast as whoever happens to check Slack. For those running agents that need sign-off outside working hours: how does the request actually reach a person, and what's your rule when the first person doesn't answer? Do you have a backup approver, or does it just wait till morning?
document fetched 2026-11-14 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xux90c/agents_that_ask_for_approval_while_youre_away