Skip to content

README.md

Latest commit

 

History

History
37 lines (21 loc) · 2.93 KB

README.md

File metadata and controls

37 lines (21 loc) · 2.93 KB
Raw

agent-framework-playground

Learning playground for Microsoft Agent Framework (MAF) - exploring workflow concepts with practical examples.

Prerequisites

  • Python 3.13+
  • uv package manager

Installation

uv sync

Examples

  • Built-in Agent (built_in_agent.py) - Demonstrates using the built-in ChatAgent executor with Azure OpenAI for creating AI agents in workflows. Shows a writer-reviewer pipeline with streaming execution. Documentation

  • Direct Edges (direct_edge.py) - Chain operations that always execute in sequence. Documentation

  • Conditional Edges (conditional_edge.py) - Branch execution based on runtime conditions. Documentation

  • Switch-Case Edges (switch_edge_cases.py) - Multi-way branching with case conditions and default fallback. Documentation

  • Fan-Out Edges (Broadcast) (fan_out_edges_all.py) - Distribute messages to multiple executors in parallel, where all targets receive every message. Documentation

  • Fan-Out Edges (Selective) (fan_out_edges_selective.py) - Selectively route messages to specific executors based on custom logic using selection_func. Note: Requires a newer library version than currently published (>1.0.0b251028). Documentation

  • Fan-In Edges (fan_out_edges_all_fan_in.py) - Collect messages from multiple executors into a single aggregator, demonstrating the scatter-gather pattern. Documentation

  • Streaming Workflow Execution (fan_out_edges_all_streaming.py) - Process workflow events in real-time using run_stream() instead of waiting for completion. Note: Currently does not work with available library version (1.0.0b251028). Documentation

Resources