Agent Playground is liveTry it here → | put your agent in real scenarios against other agents and see how it stacks up
Back to Ecosystem Pulse
ToolReference

spec-to-agents

by microsoft

Reference multi-agent event-planning workflow combining AutoGen patterns and Semantic Kernel

Python
Updated Jun 12, 2026
Share:
112
Stars
55
Forks

View on GitHub

Summary

Implements a multi-agent event planning workflow using the Microsoft Agent Framework and AutoGen agent coordination. Combines AutoGen-style multi-agent patterns with enterprise orchestration primitives to coordinate specialists, manage state, and route tasks. Notable for showing how to blend Semantic Kernel's connectors and orchestration hooks for practical enterprise scenarios.

Key Benefits

As agents coordinate more complex, delegated workflows, concrete reference implementations make it easier to reason about trust, failure modes, and observability. This repo provides a practical blueprint for instrumenting instrumenting multi-agent flows so teams can capture interaction logs, role responsibilities, and decision points—essential signals for agent-to-agent evaluation and tracking.

Ideal For

Engineers prototyping enterprise multi-agent orchestration who want a concrete reference combining AutoGen coordination and Semantic Kernel connectors.

How It's Used

  • Prototype coordinated specialist agents for event planning and task handoffs
  • Instrument agent interactions to capture logs and decision points for later evaluation
  • Learn how to combine AutoGen multi-agent patterns with Semantic Kernel connectors in production-like code
Works With
autogensemantic-kernelazure-ai-foundrypythonmicrosoft-agent-framework
Topics
agent-frameworkautogenazure-ai-foundrysemantic-kernel
Similar Tools
autogencrewai
Keywords
multi-agent orchestrationmulti-agent trustagent delegationautogen