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
ProtocolReferenceMCP

samples

by strands-agents

Runnable Strands Agents SDK examples for multi-agent flows and telemetry

Python
Updated Jul 17, 2026
Share:
810
Stars
425
Forks

View on GitHub

Summary

Showcases working agent patterns and demos built with the Strands Agents SDK. Provides runnable examples that wire up connectors (OpenAI, Anthropic, Bedrock, local LLMs), multi-agent flows, and telemetry hooks so you can replicate real agent interactions. Includes concrete patterns for delegation, message routing, and observability to help teams bootstrap multi-agent behaviors. For scalable coordination, consider the Orchestrator-Worker Pattern as a structural model, and you can also explore Human-in-the-Loop to incorporate validation and feedback loops.

Key Benefits

As multi-agent systems proliferate, practitioners need concrete examples to reproduce, test, and evaluate agent interactions rather than just API docs. These samples make it easier to capture agent interaction logs, simulate delegation patterns, and inspect behavior that feeds into A2A evaluation and reputation pipelines. Until you can run repeatable scenarios, building trust signals and agent track records is hard—these examples lower that barrier. See the Agent Registry Pattern for mechanisms that catalog and reference agent capabilities across runs.

Ideal For

Developers prototyping multi-agent orchestration and teams who want runnable examples to instrument agent interactions for evaluation and monitoring. This material aligns with patterns like the Hierarchical Multi-Agent Pattern to structure complex agent trees and workflows.

Use Cases

  • Prototype multi-agent delegation flows to reproduce failure modes and design evaluation scenarios
  • Instrument agent interactions with telemetry to feed A2A evaluation or reputation systems
  • Bootstrap integrations with OpenAI, Anthropic, Bedrock, and local LLMs to compare agent behavior
  • Create repeatable sample scenarios for pre-production agent testing and continuous agent evaluation
Works With
openaianthropic
Topics
agenticagentic-aiagentsaianthropicautonomous-agentsbedrockgenailitellmllama+9 more
Similar Tools
autogencrewai
Keywords
multi-agent trustagent-to-agent evaluationmulti-agent orchestrationagent track record