neuron-streaming

Neuron AI Streaming

This skill covers how output leaves a running agent or workflow and reaches a consumer: the native chunk stream, portable progress events, protocol adapters for UI libraries, and channels for push delivery.

The Mental Model

A Workflow segment produces a sequence of live objects while it runs. Four independent concerns decide what happens to them:

Adapter and channel compose. The adapter decides the shape, the channel the destination. The Workflow converts each item once into a ProtocolEvent. With both an adapter and channel configured, it consumes the pipeline eagerly and delivers events to the channel; otherwise the caller consumes a lazy generator. It never frames bytes: the edge that owns the transport encodes it.

Live output is ephemeral. Nothing yielded during a segment is stored in persistence or replayed when a completed step is restored. Chat history is the record the UI reconciles from. Never make correctness depend on a client receiving a streamed item.

Pull Streaming: Native Chunks

Agent::stream() and Workflow::events() return a lazy Generator unless both an adapter and channel are configured. Iterate that generator for live output, then read the final state from getReturn(). With adapter + channel, they stream eagerly to the channel and return the final AgentState or WorkflowState directly.

use NeuronAI\Chat\Messages\Stream\Chunks\TextChunk;
use NeuronAI\Chat\Messages\UserMessage;

$stream = $agent->stream(new UserMessage('Hello'));

foreach ($stream as $item) {
    if ($item instanceof TextChunk) {
        echo $item->content;
    }
}

$state = $stream->getReturn();
$message = $state->getMessage();