This week's AI automation and MCP coverage was small in scale but practical. Three articles stood out, each showing a different side of agents at work: a consumer app that handles bookings, a measure of real commercial demand, and a developer tooling update.
The Zest concierge shows agents taking over a full task, from recommendation to reservation. DoorDash's CEO said third-party AI agents have placed only several hundred orders, against about 3 million daily deliveries, which keeps the trend in proportion. The Datasette update is the most modest of the three, but it removes a setup step for users.
Zest's AI Restaurant Concierge Now Books Tables Over Text
Zest, a restaurant discovery app launched earlier this year, added a text-message concierge that recommends restaurants and answers questions about them. Users can text it directly or add it to a group chat to vote on where to eat. It books tables through OpenTable, Resy or SevenRooms, according to TechCrunch.
Best for: Groups planning dinner who want to decide and book in one chat
DoorDash says only several hundred orders came from AI agents
DoorDash CEO Tony Xu said on the third-quarter earnings call that third-party AI agents, such as Muse, Instinct and Fo, have placed several hundred orders. Set against about 3 million deliveries on an average day, the figure is small. Xu said most users still order through DoorDash's own app and that the company does not want to ignore or overstate the trend.
Best for: Readers gauging how much volume AI agents bring to commerce
datasette-llm 0.1a9 Now Uses LLM's Default Model
Simon Willison released datasette-llm 0.1a9, a small integration plugin that other Datasette plugins depend on. It now reads the default model configured in LLM, so Datasette Agent can run without a separate default-model setting. According to the post, users previously had to pass a model through an option. Willison describes it as a tiny user experience improvement rather than a new capability.
Best for: Developers maintaining Datasette plugins or running Datasette Agent locally
The takeaway
For teams building automation, the practical step is to check which defaults and integrations your tools already share before adding more configuration. Treat agent demand figures such as DoorDash's as a baseline rather than a forecast, and test booking or ordering flows against the platforms you actually use.
Comments (0)