Beyond the Joke: What is dd-cli?

While sudo make me a sandwich jokes about automation, DoorDash CLI—dd-cli—lets developers embed food and grocery ordering into their workflows or AI agents.

The tool is on a waitlist for macOS developers in the U.S. and Canada. It gives programmatic access to DoorDash: search stores, surface local deals, and run the full checkout. In other words, a consumer service becomes a scriptable utility.

The Rise of Agentic Commerce

Exposing the platform via a command line hands raw building blocks to AI agents that can act as autonomous shoppers.

Instead of a human clicking through a GUI, an AI can parse JSON, read complex menus, and fire Python scripts to place an order. Imagine software ordering lunch when a calendar shows a gap or restocking groceries as sensor data changes.

DoorDash has already tried iMessage integration, its "Ask DoorDash" chatbot, and limited exposure to LLMs like OpenAI’s ChatGPT and Anthropic’s Claude.

Enabling Developer-Led Innovation

Moving to a CLI invites developers to craft hyper-specific tools. A team could build a Slack bot that orders lunch for engineers, a pantry-management system, or a smart-home routine that orders when fridge inventory runs low.

The beta sign-up asks developers what they plan to build, signaling DoorDash’s hunt for high-utility use cases beyond simple automation.

Key Takeaways

  • Programmable Logistics: dd-cli lets you search, find deals, and checkout from the terminal or via AI agents.
  • Agentic Commerce Shift: DoorDash moves from a consumer app to a component of the agentic AI ecosystem.
  • Developer-First Approach: The macOS beta targets U.S. and Canadian developers building autonomous, software-driven commerce flows.

From a joke to a real integration

The sudo make me a sandwich gag has long mocked automation. dd-cli gives that joke real utility. It exposes DoorDash’s core functions—store search, deal surfacing, checkout—through commands you can script or call from other programs. By delivering these capabilities in a terminal, DoorDash becomes an API developers can embed anywhere.

The rise of “agentic commerce”

“Agentic commerce” means software agents act as independent shoppers. With dd-cli, an AI can read a JSON menu, check price or diet constraints, and place a purchase without a human click. The same logic could trigger a lunch order when a calendar shows free time or restock pantry items when sensors report low levels. DoorDash has already dabbled in conversational interfaces—iMessage, its own chatbot, and limited LLM exposure. dd-cli is the next step: giving agents raw, unmediated access to the ordering pipeline.

What developers can actually build

The beta asks participants to describe their use case, showing DoorDash wants high-impact integrations.

Potential projects:

  • A Slack bot that lets an engineering team order lunch with a single slash command and automatically splits the bill.
  • A smart-home routine that checks fridge sensors and orders groceries when staples run low.
  • A CI/CD pipeline that orders a celebratory pizza after a successful deployment.

Because the interface lives in the shell, these tools can be combined with existing automation frameworks, version-controlled, and audited like any other code.

Counter-arguments and concerns

Not everyone calls this a win. Giving code the power to place orders raises security questions: a compromised script could generate unwanted charges or leak payment details.

What to watch next

Takeaway

dd-cli turns a manual, app-driven experience into a programmable service that AI agents can use as a building block. Whether it stays a niche convenience or becomes a cornerstone of “agentic” ordering will depend on how fast it expands, how safely it’s managed, and what creative solutions developers unleash from the terminal.