ArchitectureA supervised JVM-class runtime — OLTP on seven engines, OLAP on three. AI-native, MCP-native, observable as plain SQL.Read the architecture
Está viendo la edición Perú. Está viendo la edición Colombia. You're viewing the Pakistan edition. Cambiar a la edición global →Cambiar a la edición global →Switch to the global edition →

Agents reach the web: provider-native search, plus HTTP and search tools

OpenAI, Anthropic and Vertex AI web search are switched on through each provider's own mechanism. An HTTP tool gives agents bounded outbound calls, and a search tool returns titles, addresses and snippets.

An agent that cannot leave the platform cannot check a supplier's site or read a regulation. This release opens the door in two ways, each bounded.

Provider-native search

  • OpenAI The chat model accepts web search options.
  • Anthropic The built-in web search tool is enabled on request, and the server-side tool use and result blocks are handled in the reply.
  • Vertex AI Google Search grounding through a search tool on the model.

Platform tools

  • HTTP GET, POST, PUT, PATCH and DELETE with headers, body and timeout, an optional host restriction, a cap on response size and a maximum number of redirects, loaded as a toolkit.
  • Search A tool that queries a public search engine and returns structured titles, addresses and snippets, with unit and live tests.

Both tools pass the same approval gate as every other. Reaching the web is a grant, not a default.