azure-ai-agentserver-responses

What it is and what it does

This package extends azure-ai-agentserver-core by adding a complete Responses protocol implementation for Azure AI Hosted Agent containers. It handles the full lifecycle of agent responses—creation, streaming via Server-Sent Events (SSE), cancellation, deletion, and replay—through REST endpoints and async handler decorators.

Developers register async handlers using the @app.response_handler decorator to process incoming requests. The package manages response state transitions (created → in_progress → completed/failed/cancelled), handles client disconnections and cancellation signals independently from server shutdown, and provides utilities like TextResponse for simple text replies and ResponseEventStream for complex multi-type outputs with function calls and reasoning items. It supports streaming, background execution, and opt-in crash recovery for stored responses.

Use it for

• Building conversational AI agents that stream token-by-token responses in real-time to clients over SSE.
• Implementing multi-user agent sessions where per-request call IDs route tool invocations to user-scoped results.
• Creating agents that emit function calls, reasoning traces, and structured outputs alongside text in a single response stream.
• Handling graceful cancellation and server shutdown independently—observing client cancellation via a signal while tracking server drain separately.
• Recovering from crashes in background response handlers by replaying persisted stream events when resilient_background mode is enabled.

Worth the install?

AI-flagged interpretation of the facts on this page. Verify before relying on it.

Install

azure-ai-agentserver-responses on PyPI

Before you install

Low friction installation with a pure-Python wheel. Actively maintained with a recent release (3 days old) and strong repository signals (5588 stars, last commit 2026-08-14). Requires Python 3.10 or later.

Python 3.10 or later required; handlers must be async def with exactly three positional parameters (request, context, cancellation_signal).

License in practice

MIT license permits free use, modification, and distribution with minimal restrictions—suitable for both open-source and commercial projects.

Quickstart

pip install azure-ai-agentserver-responses

from azure.ai.agentserver.responses import ResponsesAgentServerHost, TextResponse, ResponseContext, CreateResponse
import asyncio

app = ResponsesAgentServerHost()

@app.response_handler
async def handler(request: CreateResponse, context: ResponseContext, cancellation_signal: asyncio.Event):
    text = await context.get_input_text()
    return TextResponse(context, request, text=f"Echo: {text}")

app.run()

Package facts

Evidence: azure_ai_agentserver_responses-2.0.0-py3-none-any.whl

Tags

More Application Frameworks packages

See also azure-ai-agentserver-core · azure-ai-agentserver-invocations · clarifai-protocol · azure-ai-agentserver-agentframework · agent-framework-azurefunctions · agent-framework-openai · azure-ai-projects · agent-framework-azure-ai · agent-framework-azure-ai-search · agent-framework-foundry-hosting

Further reading