Book flight, dinner and train with an agent (failed)
Poster gave an agent a flight, a London dinner for two and a train ticket; it finished none, with OpenTable blocking robots. Task: Suggested implementation: An agent uses a headless browser to attempt booking a Saturday 8pm London dinner and a King's Cross–Edinburgh train. The workflow involves checking robots.txt, identifying itself in the User-Agent, and avoiding CAPTCHAs or personal data entry. Human role: Monitor for access blocks and decide whether to switch to approved APIs.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites: Access to approved booking APIs is likely required, as direct web scraping is often blocked by robots.txt and bot detection. Operating cost: Low for API calls, high for failed scraping attempts. Integration uncertainty: High, as many consumer booking sites restrict automated access. Human approval needed for any data entry..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
Poster gave an agent a flight, a London dinner for two and a train ticket; it finished none, with OpenTable blocking robots.
Workflow
Suggested implementation: An agent uses a headless browser to attempt booking a Saturday 8pm London dinner and a King's Cross–Edinburgh train. The workflow involves checking robots.txt, identifying itself in the User-Agent, and avoiding CAPTCHAs or personal data entry. Human role: Monitor for access blocks and decide whether to switch to approved APIs.
Reported result
Outcome unknown due to missing source evidence. Proposed acceptance check: Verify if the agent can retrieve availability data from at least one compliant API without triggering bot blocks. If no successful execution is documented, assume failure in direct web scraping due to robots.txt restrictions.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites: Access to approved booking APIs is likely required, as direct web scraping is often blocked by robots.txt and bot detection. Operating cost: Low for API calls, high for failed scraping attempts. Integration uncertainty: High, as many consumer booking sites restrict automated access. Human approval needed for any data entry.
Try it
Act as an automation engineer. I want to test an agent's ability to check availability for a Saturday 8pm dinner in London and a train from King's Cross to Edinburgh. Use a headless browser with a custom User-Agent that identifies the bot. Strictly check robots.txt before each request; do not bypass CAPTCHAs or enter personal details. First, perform a dry run on [replaceable input: specific restaurant or train API endpoint] to verify access. If blocked, report the failure reason. If successful, output a JSON with {status, availability_data, error_code}. Do not proceed to booking without my explicit confirmation. Ask clarifying questions if the target API or website is not specified.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.