Last updated:
Quick answer
Decide in advance who approves what while you are away, what simply waits, and what the customer hears in the meantime. Put it in the agent's instructions, keep money and refunds with a named person, and have the agent send a holding message instead of going silent.
Free PDF
Get this setup as a PDF
Confirm your email and we'll open a print-ready copy of this setup you can save as a PDF. You also get the free starter pack and one setup a week. Unsubscribe anytime.
Already a subscriber? Unlock it on this device
At a glance
All outcome figures are estimates from cited sources, not independent verification by us.
- What it is allowed to do
- Read-only
Only reads, researches, or summarizes. It never drafts outbound content or changes anything.
- Time saved (estimate)
- About 20 minutes before each time away.
- Difficulty
- Beginner
- Tools needed
- ChatGPT, Claude
- Typical cost
- Free tiers can draft the plan. Prices: Not stated in sources.
Reviewed Nov 2, 2026
Basis (estimate): Estimate only: based on the steps above, not timed.
What does this setup do?
Write an away plan for your AI agent: who approves which actions while you are out, what waits, and the holding message customers get so nothing stalls silently.
Who is it for?
Owners who approve their agent's drafts and worry about refunds, quotes or replies stalling when they are on a job, sick or on vacation.
How do you set it up?
Tools
- ChatGPT or Claude
- The agent's standing instructions
- A trusted backup person, if you have one
Steps
- List what normally waits for your approval: replies, quotes, refunds, bookings.
- For each, pick one: a named backup may approve, it waits until you are back, or the customer gets a holding message.
- Keep money decisions with a named person. Anthropic says its Google Workspace connectors ask before sending by default; keep that on while you are away too.
- Use the prompt below to turn your choices into an away plan, and paste it into the agent's instructions before you leave.
- Test it: send yourself a fake refund request and check the agent sends the holding message, not a decision.
- When you are back, ask the agent for a list of everything that waited.
Example output
Illustrative. Your results will differ.
- Away Nov 10 to 14. Backup: Sam (office) approves replies and bookings.
- Refunds: wait for owner; send holding message.
- Holding message: Thanks, we got your request. The owner is back Nov 15 and will reply by Nov 16.
Where to paste this prompt
ChatGPT → Project instructions, or ChatGPT Tasks for scheduled reminders (once a day on Free)
Copy-paste prompt
Adapt bracketed fields to your situation.
Write an away plan for my AI agent. Use only my choices.
I am away: [DATES]
Backup person (or none): [NAME, ROLE]
For each item, my choice (backup approves / waits for me / holding message):
- Customer replies: [CHOICE]
- Quotes: [CHOICE]
- Refunds and credits: [CHOICE]
- Bookings and changes: [CHOICE]
Include:
1. A short rule list the agent follows while I am away
2. A holding message under 50 words that tells the customer when they will hear back, with no promises about the outcome
3. A rule: anything not listed waits for me
Do not give the backup any permission I did not list.Newsletter
Get the free starter pack
Free starter pack with five beginner copy-paste prompts after you confirm, plus one researched setup in your inbox each week. Unsubscribe anytime.
What usually goes wrong?
Where agents tend to stall on this job. More patterns in Where AI gets stuck.
- Letting the agent decide refunds because no one is around.
- Promising a resolution date the backup cannot keep.
- Forgetting to switch the plan off when you return.
Which chat tools work with this prompt?
Paste the prompt into any of these assistants (free or paid tiers).
- ChatGPT
- Claude
Common questions
- What should my AI agent do when I am not around to approve?
- Follow an away plan: named backup approvals for some items, waiting for others, and a holding message to the customer.
- Should the agent approve refunds itself?
- No. Keep money decisions with a named person and have the agent send a holding message instead.
- How do I test it?
- Send yourself a fake request before you leave and check the agent follows the plan.
What tips make it smoother?
- A holding message beats silence; customers mostly want to know someone saw it.
- Give the backup a short list, not your whole inbox.
- Set an end date so the plan does not linger.
Where do the claims come from?
Summaries use public sources linked below, not independent verification. How setups are picked.
- Read Anthropic’s write-up (support.claude.com)
Anthropic: Use Google Workspace connectors: approval before send, reply or forward by default (read 2026-10-31 through a page fetch tool, not a direct browser open)
- Read OpenAI’s write-up (help.openai.com)
OpenAI: ChatGPT agent: when signed into sites or apps it can access sensitive data and perform actions (read 2026-11-02 from a search result excerpt of the page)
Related setups
- Make your AI agent stop for approval
Make your AI agent stop and show you every email, booking or post as a draft with a one-line summary, and wait for your OK before it goes out.
- Set guardrails before an AI agent gets access
Write a one-page rulebook for an AI agent before it touches your accounts: what it can read, what it can only draft, what always needs your approval, and what it must never do.
- Reply to an unreasonable refund request
Draft a calm, firm reply to a refund demand you should not grant, paired with an alternative that fits your policy.