Stream the work

Iterate the run to get events as data — route them to a GUI, a websocket, metrics, or break mid-stream to cancel. Each iteration yields a typed event: ToolCall (a tool-call boundary, finished=False on start and True on finish) or TextChunk (the answer, as it streams).

import sys
from cosmon_agent_sdk import Client, TextChunk, ToolCall

with Client() as client:
    for event in client.run("Mesh and solve the model", agent="abaqus"):
        if isinstance(event, TextChunk):
            print(event.text, end="", flush=True)          # answer, token by token
        elif isinstance(event, ToolCall) and not event.finished:
            print(f"▶ {event.summary}", file=sys.stderr)    # tool steps

A run is single-use: iterate it or call .text(), not both. The TextChunks reconstruct the same answer .text() returns.

Guide: streaming the run

Observe vs respond, agent attribution, and cancelling mid-stream.

One-shot run Attach files

Getting Started

CAD

CAE

Agent SDK

Deployment

Guide: streaming the run