[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-71954":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":23,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":35,"readmeContent":36,"aiSummary":37,"trendingCount":16,"starSnapshotCount":16,"syncStatus":38,"lastSyncTime":39,"discoverSource":40},71954,"agent-framework","microsoft\u002Fagent-framework","microsoft","A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.","https:\u002F\u002Faka.ms\u002Fagent-framework",null,"Python",11226,1879,101,496,0,137,315,874,411,44.82,"MIT License",false,"main",[5,26,27,28,29,30,31,32,33,34],"agentic-ai","agents","ai","dotnet","multi-agent","orchestration","python","sdk","workflows","2026-06-12 02:02:56","![Microsoft Agent Framework](docs\u002Fassets\u002Freadme-banner.png)\n\n# Welcome to Microsoft Agent Framework!\n\n[![Microsoft Foundry Discord](https:\u002F\u002Fdcbadge.limes.pink\u002Fapi\u002Fserver\u002Fb5zjErwbQM?style=flat)](https:\u002F\u002Fdiscord.gg\u002Fb5zjErwbQM)\n[![MS Learn Documentation](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FMS%20Learn-Documentation-blue)](https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fagent-framework\u002F)\n[![PyPI](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fagent-framework)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fagent-framework\u002F)\n[![NuGet](https:\u002F\u002Fimg.shields.io\u002Fnuget\u002Fv\u002FMicrosoft.Agents.AI)](https:\u002F\u002Fwww.nuget.org\u002Fprofiles\u002FMicrosoftAgentFramework\u002F)\n[![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002Fagent-framework?style=social)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fstargazers)\n\n\nMicrosoft Agent Framework (MAF) is an open, multi-language framework for building **production-grade AI agents and multi-agent workflows** in **.NET and Python**.\n\nMicrosoft Agent Framework is built for teams taking agents from prototype to production. It provides a consistent foundation for building, orchestrating, and operating agent systems across Python and .NET, while keeping architecture choices open as requirements evolve, and supports a broad ecosystem including Microsoft Foundry, Azure OpenAI, OpenAI, and the GitHub Copilot SDK, with samples and hosting patterns for both local development and cloud deployment.\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=AAgdMhftj8w\" title=\"Watch the full Agent Framework introduction (30 min)\">\n    \u003Cimg src=\"https:\u002F\u002Fimg.youtube.com\u002Fvi\u002FAAgdMhftj8w\u002Fhqdefault.jpg\"\n         alt=\"Watch the full Agent Framework introduction (30 min)\" width=\"480\">\n  \u003C\u002Fa>\n\u003C\u002Fp>\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=AAgdMhftj8w\">\n    Watch the full Agent Framework introduction (30 min)\n  \u003C\u002Fa>\n\u003C\u002Fp>\n\n## Is this the right framework for you?\n\nMAF is a strong fit if you:\n- are building agents and workflows you expect to run in production,\n- need orchestration beyond a single prompt or stateless chat loop,\n- want graph-based patterns such as sequential, concurrent, handoff, and group collaboration,\n- care about durability, restartability, observability, governance, or human-in-the-loop control,\n- need provider flexibility so your architecture can evolve without major rewrites.\n\n## Key Features\nExplore new MAF capabilities and real implementation patterns on the [official blog](https:\u002F\u002Fdevblogs.microsoft.com\u002Fagent-framework\u002F).\n\n- **Python and C#\u002F.NET Support**: Full framework support for both Python and C#\u002F.NET implementations with consistent APIs\n  - [Python packages](.\u002Fpython\u002Fpackages\u002F) | [.NET source](.\u002Fdotnet\u002Fsrc\u002F)\n- **Multiple Agent Provider Support**: Support for various LLM providers with more being added continuously\n  - [Python examples](.\u002Fpython\u002Fsamples\u002F02-agents\u002Fproviders\u002F) | [.NET examples](.\u002Fdotnet\u002Fsamples\u002F02-agents\u002FAgentProviders\u002F)\n- **Middleware**: Flexible middleware system for request\u002Fresponse processing, exception handling, and custom pipelines\n  - [Python middleware](.\u002Fpython\u002Fsamples\u002F02-agents\u002Fmiddleware\u002F) | [.NET middleware](.\u002Fdotnet\u002Fsamples\u002F02-agents\u002FAgents\u002FAgent_Step11_Middleware\u002F)\n- **Orchestration Patterns & Workflows**: Build multi-agent systems with graph-based workflows supporting sequential, concurrent, handoff, and group collaboration patterns; includes checkpointing, streaming, human-in-the-loop, and time-travel\n  - [Python workflows](.\u002Fpython\u002Fsamples\u002F03-workflows\u002F) | [.NET workflows](.\u002Fdotnet\u002Fsamples\u002F03-workflows\u002F)\n- **Foundry Hosted Agents (new)**: Deploy and host your agents to Foundry-hosted infrastructure with just 2 additional lines of code\n  - [Python samples](.\u002Fpython\u002Fsamples\u002F04-hosting\u002Ffoundry-hosted-agents\u002F) | [.NET samples](.\u002Fdotnet\u002Fsamples\u002F04-hosting\u002FFoundryHostedAgents\u002F)\n- **Observability**: Built-in OpenTelemetry integration for distributed tracing, monitoring, and debugging\n  - [Python observability](.\u002Fpython\u002Fsamples\u002F02-agents\u002Fobservability\u002F) | [.NET telemetry](.\u002Fdotnet\u002Fsamples\u002F02-agents\u002FAgentOpenTelemetry\u002F)\n- **Declarative Agents**: Define agents using YAML for faster setup and versioning\n  - [Declarative agent samples](.\u002Fdeclarative-agents\u002F)\n- **Agent Skills**: Build domain-specific knowledge bases from multiple sources—files, inline code, class libraries—for agents to discover and use\n  - [Skills design](.\u002Fdocs\u002Fdecisions\u002F0021-agent-skills-design.md)\n- **AF Labs**: Experimental packages for cutting-edge features including benchmarking, reinforcement learning, and research initiatives\n  - [Labs directory](.\u002Fpython\u002Fpackages\u002Flab\u002F)\n- **DevUI**: Interactive developer UI for agent development, testing, and debugging workflows\n  - [See the DevUI in action](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=mOAaGY4WPvc)\n\n## Table of Contents\n\n- [Getting Started](#getting-started)\n  - [Installation](#installation)\n  - [Learning Resources](#learning-resources)\n  - [Quickstart](#quickstart)\n    - [Basic Agent - Python](#basic-agent---python)\n    - [Basic Agent - .NET](#basic-agent---net)\n- [More Examples & Samples](#more-examples--samples)\n- [Community & Feedback](#community--feedback)\n- [Troubleshooting](#troubleshooting)\n- [Contributor Resources](#contributor-resources)\n\n## Getting Started\n### Installation\nPython\n\n```bash\npip install agent-framework\n# This will install all sub-packages, see `python\u002Fpackages` for individual packages.\n# It may take a minute on first install on Windows.\n```\n\n.NET\n\n```bash\ndotnet add package Microsoft.Agents.AI\n# For Foundry integration (used in the .NET quickstart below):\ndotnet add package Microsoft.Agents.AI.Foundry\ndotnet add package Azure.AI.Projects\ndotnet add package Azure.Identity\n```\n\n### Learning Resources\n\n- **[Overview](https:\u002F\u002Flearn.microsoft.com\u002Fagent-framework\u002Foverview\u002Fagent-framework-overview)** - High level overview of the framework\n- **[Quick Start](https:\u002F\u002Flearn.microsoft.com\u002Fagent-framework\u002Ftutorials\u002Fquick-start)** - Get started with a simple agent\n- **[Tutorials](https:\u002F\u002Flearn.microsoft.com\u002Fagent-framework\u002Ftutorials\u002Foverview)** - Step by step tutorials\n- **[User Guide](https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fagent-framework\u002Fuser-guide\u002Foverview)** - In-depth user guide for building agents and workflows\n- **[Migration from Semantic Kernel](https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fagent-framework\u002Fmigration-guide\u002Ffrom-semantic-kernel)** - Guide to migrate from Semantic Kernel\n- **[Migration from AutoGen](https:\u002F\u002Flearn.microsoft.com\u002Fen-us\u002Fagent-framework\u002Fmigration-guide\u002Ffrom-autogen)** - Guide to migrate from AutoGen\n\n### Quickstart\n\n#### Basic Agent - Python\n\nCreate a simple Azure Responses Agent that writes a haiku about the Microsoft Agent Framework\n\n```python\n# pip install agent-framework\n# Use `az login` to authenticate with Azure CLI\nimport os\nimport asyncio\nfrom agent_framework import Agent\nfrom agent_framework.foundry import FoundryChatClient\nfrom azure.identity import AzureCliCredential\n\n\nasync def main():\n    # Initialize a chat agent with Microsoft Foundry\n    # the endpoint, deployment name, and api version can be set via environment variables\n    # or they can be passed in directly to the FoundryChatClient constructor\n    agent = Agent(\n      client=FoundryChatClient(\n          credential=AzureCliCredential(),\n          # project_endpoint=os.environ[\"FOUNDRY_PROJECT_ENDPOINT\"],\n          # model=os.environ[\"FOUNDRY_MODEL_DEPLOYMENT_NAME\"],\n      ),\n      name=\"HaikuAgent\",\n      instructions=\"You are an upbeat assistant that writes beautifully.\",\n    )\n\n    print(await agent.run(\"Write a haiku about Microsoft Agent Framework.\"))\n\nif __name__ == \"__main__\":\n    asyncio.run(main())\n```\n\n#### Basic Agent - .NET\nCreate a simple Agent, using Microsoft Foundry that writes a haiku about the Microsoft Agent Framework\n\n```c#\n\u002F\u002F This sample shows how to create and run a basic agent with AIProjectClient.AsAIAgent(...).\n\nusing Azure.AI.Projects;\nusing Azure.Identity;\nusing Microsoft.Agents.AI;\n\nstring endpoint = Environment.GetEnvironmentVariable(\"AZURE_AI_PROJECT_ENDPOINT\") ?? throw new InvalidOperationException(\"AZURE_AI_PROJECT_ENDPOINT is not set.\");\nstring deploymentName = Environment.GetEnvironmentVariable(\"AZURE_AI_MODEL_DEPLOYMENT_NAME\") ?? \"gpt-5.4-mini\";\n\nAIAgent agent =\n    new AIProjectClient(new Uri(endpoint), new DefaultAzureCredential())\n    .AsAIAgent(model: deploymentName, instructions: \"You are an upbeat assistant that writes beautifully.\", name: \"HaikuAgent\");\n\n\u002F\u002F Once you have the agent, you can invoke it like any other AIAgent.\nConsole.WriteLine(await agent.RunAsync(\"Write a haiku about Microsoft Agent Framework.\"));\n```\n\n## More Examples & Samples\n\n### Python\n\n- [Getting Started](.\u002Fpython\u002Fsamples\u002F01-get-started): progressive tutorial from hello-world to hosting\n- [Agent Concepts](.\u002Fpython\u002Fsamples\u002F02-agents): deep-dive samples by topic (tools, middleware, providers, etc.)\n- [Workflows](.\u002Fpython\u002Fsamples\u002F03-workflows): workflow creation and integration with agents\n- [Hosting](.\u002Fpython\u002Fsamples\u002F04-hosting): A2A, Azure Functions, Durable Task hosting\n- [End-to-End](.\u002Fpython\u002Fsamples\u002F05-end-to-end): full applications, evaluation, and demos\n\n### .NET\n\n- [Getting Started](.\u002Fdotnet\u002Fsamples\u002F01-get-started): progressive tutorial from hello agent to hosting\n- [Agent Concepts](.\u002Fdotnet\u002Fsamples\u002F02-agents\u002FAgents): basic agent creation and tool usage\n- [Agent Providers](.\u002Fdotnet\u002Fsamples\u002F02-agents\u002FAgentProviders): samples showing different agent providers\n- [Workflows](.\u002Fdotnet\u002Fsamples\u002F03-workflows): advanced multi-agent patterns and workflow orchestration\n- [Hosting](.\u002Fdotnet\u002Fsamples\u002F04-hosting): A2A, Durable Agents, Durable Workflows\n- [End-to-End](.\u002Fdotnet\u002Fsamples\u002F05-end-to-end): full applications and demos\n\n## Community & Feedback\n\n- **Found a bug?** File a [GitHub issue](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fissues) to help us improve.\n- **Enjoying MAF?** [![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FStar-us%20on%20GitHub-yellow)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework) to show your support and help others discover the project.\n- **Have questions?** Join our [Discord](https:\u002F\u002Fdiscord.gg\u002Fb5zjErwbQM) or visit [weekly office hours](.\u002FCOMMUNITY.md#public-community-office-hours).\n\n## Troubleshooting\n\n### Authentication\n\n| Problem | Cause | Fix |\n|---------|-------|-----|\n| Authentication errors when using Azure credentials | Not signed in to Azure CLI | Run `az login` before starting your app |\n| API key errors | Wrong or missing API key | Verify the key and ensure it's for the correct resource\u002Fprovider |\n\n> **Tip:** `DefaultAzureCredential` is convenient for development but in production, consider using a specific credential (e.g., `ManagedIdentityCredential`) to avoid latency issues, unintended credential probing, and potential security risks from fallback mechanisms.\n\n### Environment Variables\nFor environment variable configuration specific to each sample, refer to the README in the sample directory ([Python samples](.\u002Fpython\u002Fsamples\u002F) | [.NET samples](.\u002Fdotnet\u002Fsamples\u002F)).\n\n## Contributor Resources\n\n- [Contributing Guide](.\u002FCONTRIBUTING.md)\n- [Python Development Guide](.\u002Fpython\u002FDEV_SETUP.md)\n- [Design Documents](.\u002Fdocs\u002Fdesign)\n- [Architectural Decision Records](.\u002Fdocs\u002Fdecisions)\n\n## Important Notes\n\n> [!IMPORTANT]\n> If you use Microsoft Agent Framework to build applications that operate with any third-party servers, agents, code, or non-Azure Direct models (“Third-Party Systems”), you do so at your own risk. Third-Party Systems are Non-Microsoft Products under the Microsoft Product Terms and are governed by their own third-party license terms. You are responsible for any usage and associated costs.\n>\n>We recommend reviewing all data being shared with and received from Third-Party Systems and being cognizant of third-party practices for handling, sharing, retention and location of data. It is your responsibility to manage whether your data will flow outside of your organization’s Azure compliance and geographic boundaries and any related implications, and that appropriate permissions, boundaries and approvals are provisioned.\n> \n>You are responsible for carefully reviewing and testing applications you build using Microsoft Agent Framework in the context of your specific use cases, and making all appropriate decisions and customizations. This includes implementing your own responsible AI mitigations such as metaprompt, content filters, or other safety systems, and ensuring your applications meet appropriate quality, reliability, security, and trustworthiness standards. See also: [Transparency FAQ](.\u002FTRANSPARENCY_FAQ.md)\n","Microsoft Agent Framework (MAF) 是一个用于构建、编排和部署支持Python和.NET的生产级AI代理和多代理工作流的框架。其核心功能包括对Python和C#\u002F.NET的全面支持，提供一致的API接口；支持多种语言模型提供商，并且具有灵活的中间件系统，可用于请求\u002F响应处理、异常处理和自定义管道。此外，它还支持图基模式，如顺序、并发、交接和团队协作等复杂的工作流程。该框架适用于需要将代理从原型推进到生产的团队，特别是在需要超越单个提示或无状态聊天循环的编排、关注持久性、可恢复性、可观测性、治理或人类参与控制的场景下尤为适用。",2,"2026-06-11 03:39:41","high_star"]