Show HN: Search2o Simplifies AI Agents

Developer

User

Search2o agent serverpip install search2o

Agent framework

Build, validate, publish agents

Searchrequest →↓ agent built for it

Describe agents in EnglishTravel desk: Arranges air travel and hotel stays.

search("Book a trip to New York")

Travel desk

Agent runtime

execAgent(convId, Travel desk, "Book a trip to New York")

Booked UA1191, departs 06:00.

Built-in agent discovery and multi-agent conversations

Agents are discovered by search and build on each other in a conversation.

Travel desk UA1191 departs 06:00 Tuesday, back at 19:30, and a night at the Marriott near the office. $1,240. Book it? BookHold Submit

2Human in the loopAgent state is stored and when the user answers, it continues where it left off.

3Travel deskBooked UA1191 and the hotel night. The confirmation is in your inbox.

3Conversation contextTravel desk agent ended here. Its prompts and other information it chooses stay in the conversation. File the expenses for that trip

4ReimbursementsFiled the New York trip, $1,240, against your cost center. Sent to your manager for approval.

4Building on each otherSearch matched the follow-up request to Reimbursements. It used the saved prompt from Travel desk to do its task.

1DiscoverySearch matched the request to Travel desk agent. The UI is configured to execute the agent automatically.

2Human in the loopAgent state is stored and when the user answers, it continues where it left off.

3Conversation contextTravel desk agent ended here. Its prompts and other information it chooses stay in the conversation.

4Building on each otherSearch matched the follow-up request to Reimbursements. It used the saved prompt from Travel desk to do its task.

Built-in agent infrastructure

Agent server takes care of everything else, so your agent contains only business logic.

LLM calls Model neutrality Tool calling loop API, database and MCP calls Connection pools Secret vault Error handling Runtime limits Human in the loop Streaming Save/Restore conversations Memory Externalized prompts Draft with AI Trace and validation Agent versioning Multi‑developer publishing Live configuration Observability Audit log requestYour agentBusiness logic only

response

Multiple interfaces

A client uses just two calls to discover and run hundreds of agents.

Browser chatBundled UI or build your own.

Slack, Teams, Google ChatChat integrations

Agent harnesssearch2o-skill · Skill integration

Authorization: Bearer <token> search() → agent execAgent() → answer

Agent server Search Agent runtime

Managed state

No dependencies to install or run in your organization.

Agent servers Source available

encrypted

Search2o Cloud

State Conversations, memories encrypted by default

Built for the enterprise

Security & privacy

Secrets live in your vault. Conversations and memories are encrypted, by default or, optionally, end to end. Data privacy lists how each kind of data is handled.

Multi-user access and roles

Four roles: users, developers, administrators and owners. Service accounts for bots and scripts. Built-in authentication as well as single sign‑on.

Source available

Everything that runs in your organization can be built from the source: the agent server and the UI. Contact us for customizations for internal use.

Multiple environments

An agent refers to enterprise systems by profile name. Dev and prod environment’s profiles point at its own systems, so the same agent runs unchanged.

Horizontally scalable

Agent servers are stateless and can be run in a cluster. Configuration and profile changes are live on the next agent run on every server, without restart.

Integration

Integrate at three levels: from Python expressions inside an agent, through the REST API, and with hooks that run your code at the server's decision points.