Use ChatGPT Agent to compile used Honda Fit listings

Calvin Wankhede tested ChatGPT Agent on a used-car search that required inspecting many individual Facebook Marketplace listings. He specified Honda Fit vehicles from 2015 through 2020 within fifty kilometers of Toronto.

After he supplied a two-factor authentication code, the agent entered Marketplace, set the location and distance filter, and opened individual listings. It recorded details while he stepped away from the computer. The final report contained roughly two to three dozen matching vehicles, according to his account.

Each brief organized year and model, price, mileage, transmission and location. A notes column extracted seller details such as dealer status, salvage title or willingness to negotiate. The report also included a screenshot for each listing, giving Wankhede material to inspect when considering a specific vehicle.

The process was persistent rather than especially fast. Wankhede observed browsing loops and delayed actions that could trigger repeated back clicks, return the agent to the homepage and restart the search. In another car-search interaction, it struggled with a slider because it would not accept a fifty-one-kilometer radius instead of exactly fifty. No seller contact or vehicle purchase is reported; the result is a detailed candidate report with login assistance and navigation limitations.

Prompt
Suggested prompt (written by Agents In Use, not the author's)

Source: www.androidauthority.com · Calvin Wankhede