How to use ChatGPT agent for London Eye checkout
ChatGPT agent navigated London Eye ticket checkout for a family outing; Paul took over payment, stopped interruptions and saved confirmed tickets to his phone.
linkedin.com
Original prompt
Shared by the source author. See source notes for context.
Take me as far as possible into the checkout process
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Paul E.
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- The agent reached ticket checkout; Paul paid, received confirmation, saved the tickets and reports completing the outing.
- Human involvement
- Paul chose the attraction and ticket window, entered private details, paid with a virtual card and instructed the agent to wait during takeover.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Paul selected an outing after the agent checked availability and cross-checked website restrictions.
- The agent selected adult and child London Eye tickets and progressed through checkout extras.
- Paul took browser control for payment, corrected interruptions and saved the confirmed tickets.
Paul E. used ChatGPT agent to move a London outing with his nine-year-old niece from availability checks to ticket checkout. The afternoon had a firm end because his sister needed to start work. Initial itinerary brainstorming happened in standard ChatGPT; the agent subsequently operated websites.
When Sky Garden tickets were unavailable, the agent cross-checked the Sunday walk-in situation. Paul selected the London Eye and asked for an adult and child ticket in an afternoon time window. The agent worked through dates, ticket selection and distracting extras, then stopped for email and payment information.
Paul took over to pay using a one-time virtual card. During that handoff, the agent repeatedly asked what to do next. A further instruction to wait until he returned control stopped the interruptions. He completed payment, downloaded the tickets, saved them to his phone wallet and received the merchant confirmation.
The account describes actually attending the outing. Its useful boundary is concrete: the agent navigated the booking flow, while Paul chose the activity, supplied private information and paid. Website availability reflects that particular visit, rather than a standing rule for future travelers.
Prompt
Original prompt
Source: LinkedIn · Paul E.
Frequently asked questions
Did the agent pay for the tickets?
No. Paul took over the browser and completed payment himself.
What went wrong during the handoff?
The agent interrupted with follow-up questions until Paul explicitly told it to wait for returned control.
Was this only an itinerary exercise?
No. The author reports confirmed tickets and actually attending the outing.
More details
About this account
Paul selected an outing after the agent checked availability and cross-checked website restrictions. The agent selected adult and child London Eye tickets and progressed through checkout extras.
Original article opened in full. Original prompt excerpt is the published checkout instruction. Standard ChatGPT planning and agent browser actions are distinguished; historical venue availability is not current guidance. Original prompt excerpt.
Credit: Paul E. — LinkedIn
Source published
Added to Agents In Use
Source checked
Tools and topics
ChatGPT agent · London Eye website · Phone wallet
ticket checkout · payment takeover · family outing
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.