What it is
WebShop is a simulated shopping environment built around product information and natural-language instructions. Agents search and browse to choose a product that meets the request.
Shopping tasks combine language understanding with grounded decisions. A choice needs to satisfy the user’s product requirements.
How an agent works in it
Sees
Shopping-page content and the purchase instruction.
Does
Search, browse products and select available options.
Feedback
A task score based on how the chosen product satisfies the instruction.
Example task (illustrative): Find a product with the requested attributes and budget.
Setup
Prepare the product data, search index and environment according to the repository. Select the task split before running experiments.
Use it with BoltzLabs
BoltzLabs does not host this project. To record an agent working on one of its tasks:
- Prepare one task from the project: its starter files, instructions, dependencies and check command.
- In Evals, choose New eval, upload the starter folder or enter its repository, then set the task and check command.
- Choose BoltzLabs models or your own endpoint. Each run gets its own machine.
- When the run finishes, open it to read the agent’s trace, file changes and check result.
Projects that need a desktop or phone, self-hosted websites, a game or a simulator need that runtime set up separately; an eval machine does not include it. Machine time uses account credits; hosted model calls are billed separately.