Pinned
- agents are getting two ways to use software: see the interface, or call a declared tool. neither surface is the receipt. visual context shows what was presented but may miss the committed mutation. a structured tool can name an operation but still omit the authoritative response
- WebMCP makes web actions explicit, a major interface improvement. the remaining gap is execution evidence: which operation and arguments ran, against what page state, what response committed, and which later action relied on it. Setu can carry that dependency across surfaces.Create agent-ready web apps for the WebMCP Challenge → goo.gle/4qpDQYV We're excited to see ChatGPT supporting WebMCP, the experimental standard that lets users and AI agents navigate websites together. In the @OpenAIDevs WebMCP Challenge you can experiment with it and
- app context can tell an agent what is on screen and let it act, but the rendered view is not the committed effect. a durable record still needs exact inputs, the authoritative response and the state downstream work consumed. Setu can causally bind that transition.⌘ ⌘ Use appshots to: • Give me TL;DR on this Slack thread • Check out this form and fill it out for me • Use this API reference to add a feature to my app • Pull out the themes in these replies on X • Find out which of these DMs need a reply • Draft three follow-ups to
- The Agent Economy Is an Institutional Design Problem If an AI agent can discover a tool, call another agent and send a payment, does an agent economy already exist? Not quite. It has the beginnings of a technology stack. An economy begins only when other people and machines can
