Skip to main content

Documentation Index

Fetch the complete documentation index at: https://agent37.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

The Agents API has three resources. An instance is the agent’s computer, a session is a conversation on it, and a response is a single turn.
ConceptWhat it is
Instance inst_An isolated, always-on computer that runs your agent. Built from a template that installs one or more agents (Hermes by default). Persistent until you delete it. Create one per end user.
Session sess_One conversation on an instance, running one of its agents. A message starts one; reuse its id to continue. An instance can hold many.
Response resp_One agentic turn: your input, the agent’s work, its reply. Stream it live or fetch it by id.
The agent is not the model. The template installs the agents (Hermes, Claude Code, Codex, OpenClaw, or a mix); a session runs one of them. The model is the LLM that the agent runs on, chosen per message as model + provider. They are separate dials.

​
How they fit together

  • Create an instance once per end user. It keeps files, connected accounts, and memory across every session.
  • Start a session by sending a message with an instance_id. The reply returns a session_id you reuse to continue the thread.
  • Each message produces a response: one agentic turn you can stream live or fetch later by id.
All ids are typed and prefixed (inst_, sess_, resp_, sched_, conn_, sk_live_), field names are snake_case, and timestamps are unix seconds.

Instances

Create, size, and manage the agent’s computer.

Streaming events

Render text, reasoning, and tool activity live.

Send a message

The core call and its response shape.

Templates

Choose which agents an instance installs.