What is Search2o

Search2o is a platform for building agents and running them from a search box. A user types what they need. Search2o finds the agent built for that need, runs it, and returns the answer. The answer starts a conversation that the user can continue.

Developers write agents in JSON with Python expressions and describe each agent in plain English. Search matches a request to the agent whose description fits, and the agent runs under controlled execution.

Search2o is self-contained. Create an account and install the agent server. The search box, the editor, user management and reports come with it.

What makes it different

• Search finds the agent. An enterprise accumulates many agents, written by different teams. Users do not learn their names; they ask for what they need, and search chooses the agent. The whole collection behaves like one application.
• Agents are small and structured. An agent does one task. It is a set of functions, and a function is an ordered list of commands. Twenty commands cover what agents do: call APIs and databases, prompt LLMs with tools, ask the user for input, and remember facts across conversations.
• Agents are independent. No agent calls another. Agents build on each other's work through the conversation: a follow-up goes through search again, and the agent that runs next sees what the conversation already holds.
• Execution is controlled. The Python expressions in an agent run in the agent runtime, under the allowlist, the compile rules and the runtime limits that developers set.
• Nothing to host for state. The agent server holds no state. Everything is stored in Search2o Cloud, so agent servers can be started, stopped and scaled freely. Conversations, queries and memories are encrypted.