PitchHut logo
Smart delivery management through an AI command-line interface.
Pitch

EAT CLI offers a seamless plug-and-play solution for AI agents to browse menus, customize orders, and check out from leading delivery services. With just a single command, it integrates fully with Wolt and provides a stub for Bolt, simplifying the order process and enhancing efficiency in food delivery.

Description

EAT CLI is an innovative command-line interface that enables AI agents to seamlessly browse menus, customize orders, and complete transactions with major delivery services. Currently, EAT CLI provides full integration with Wolt and a stub integration for Bolt.

Key Features

  • Menu Browsing: Simplifies the process of searching and exploring various menu items across supported delivery platforms.
  • Order Customization: Offers flexibility to modify orders directly from the command line.
  • Easy Checkout: Facilitates a straightforward checkout process, allowing for quick order placements.

Configuration

Configuration is managed through a config.yml file, where users can specify critical parameters like the delivery service provider and success conditions. Here’s an example configuration:

provider: "wolt"
success_url_pattern: "https://wolt.com/en/discovery"
success_selector: "[data-test-id='UserStatus.ProfileImage']"
user_data_dir: "./profile/wolt"
venue_base_url: "https://wolt.com/en/lva/riga"
headless: false
timeout_seconds: 600

Command-Line Usage

EAT CLI operates with various commands, each designed to execute specific tasks:

  • Authentication Command: Create or refresh a user session to authenticate the agent:

    go run main.go auth
    
  • Search Command: Search for items based on a query and receive a structured JSON summary:

    go run main.go search <keyword>
    
  • Basket Commands: Manage the current basket state, including adding or removing items:

    go run main.go basket add <venue_slug> <item_id>
    go run main.go basket remove <venue_slug> <item_id>
    
  • Checkout Command: Place an order with the current basket's selected venue:

    go run main.go checkout <venue_slug>
    

Integration Testing

An integration harness is built into the system to validate the end-to-end functionality of searches, additions, and removals in the basket.

Extend the capabilities of your AI agents with EAT CLI, ensuring efficient order management and enhancing the overall user experience in food delivery services.

1 comment
about 21 hours ago

Thanks for featuring the project 🙏

Sign in to comment.