# microsoft/agent-framework

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/microsoft-agent-framework).**

7,277 stars · 1,182 forks · Python · mit

## Links

- GitHub: https://github.com/microsoft/agent-framework
- Homepage: https://aka.ms/agent-framework
- awesome-repositories: https://awesome-repositories.com/repository/microsoft-agent-framework.md

## Topics

`agent-framework` `agentic-ai` `agents` `ai` `dotnet` `multi-agent` `orchestration` `python` `sdk` `workflows`

## Description

The agent-framework is an LLM agent orchestration framework and multi-agent workflow engine designed for building autonomous AI agents. It provides a tool integration layer for binding external functions, APIs, and sandboxed code as executable tools for language models.

The framework distinguishes itself through a graph-based system for designing sequential and parallel task flows, featuring state management and checkpointing for long-running processes. It implements comprehensive conversational state management and an observability suite that uses telemetry to trace execution flows and monitor tool calls.

The project covers a broad range of capabilities, including retrieval augmented generation via vector database integration, human-in-the-loop approval gating for tool use, and a middleware-based request pipeline for security and telemetry. It also supports structured output enforcement, session-based context restoration, and standardized protocols for remote agent connectivity.

## Tags

### Artificial Intelligence & ML

- [Autonomous AI Agent Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/autonomous-ai-agent-frameworks.md) — Provides a comprehensive framework for building and deploying autonomous AI agents that use language models.
- [Multi-Agent Orchestration](https://awesome-repositories.com/f/artificial-intelligence-ml/multi-agent-orchestration.md) — Coordinates interactions between multiple AI entities using sequential or concurrent patterns to achieve complex goals. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/index))
- [Long-term Memory Stores](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-architectures/memory-management-systems/long-term-memory-stores.md) — Implements persistent storage mechanisms to retain user context and memories across multiple agent sessions. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/index))
- [Agent Communication Protocols](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-communication-protocols.md) — Implements common protocols to enable discovery and message exchange between agents across different frameworks. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/a2a))
- [Agent Provider Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-provider-integrations.md) — Integrates agents with various AI model providers and specialized processing services. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/agents/agent-types/index/))
- [Agent Session Management](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-session-management.md) — Saves and reloads full session state objects to resume conversations and maintain continuity across restarts. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/persisted-conversation/))
- [Agent Workflow Orchestrations](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-workflow-orchestrations.md) — Sequences specialized AI agents and functions via graph-based paths with type-safe routing. ([source](https://learn.microsoft.com/en-us/agent-framework/overview/))
- [Autonomous Agents](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents.md) — Builds autonomous entities that integrate LLMs with tool usage and planning capabilities to perform tasks. ([source](https://learn.microsoft.com/en-us/agent-framework/))
- [Agent Orchestration Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents/agent-orchestration-frameworks.md) — Acts as an orchestration framework for building autonomous agents that coordinate multi-agent workflows.
- [Agent Instantiation](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents/agent-orchestration-frameworks/agent-instantiation.md) — Enables building autonomous entities that process requests using a small set of configuration. ([source](https://learn.microsoft.com/agent-framework/tutorials/overview))
- [Agent Construction Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents/ai-agent-builders/agent-construction-frameworks.md) — Provides a framework for defining how custom data, memory, and tools are integrated into autonomous agents. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/overview))
- [Stateful Workflow Persistence](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/control-flow-and-workflows/human-in-the-loop-workflows/stateful-workflow-persistence.md) — Saves progress of long-running processes via state checkpoints to enable recovery from the last known state. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/workflows/overview/))
- [Durable Agent Runtimes](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/runtime-and-ops/durable-agent-runtimes.md) — Ensures workload durability by persisting agent and workflow state to recover long-running tasks after failures. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/hosting))
- [Multi-Agent Orchestration Patterns](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/integration-deployment/agent-frameworks/agent-orchestrators/multi-agent-orchestration-patterns.md) — Uses deterministic code-based workflows to manage complex interactions between agents in sequential or parallel patterns. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/durable-extension))
- [Agent Response Streamers](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/integration-deployment/agent-frameworks/agent-runtimes/streaming-response-processors/agent-response-streamers.md) — Ships a mechanism for streaming incremental agent outputs and tool interactions to clients in real time. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/index))
- [Agent Tool Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/integration-deployment/agent-frameworks/tool-use-and-execution/agent-tool-integrations.md) — Connects agents to external APIs, database queries, and functions to execute specific tasks automatically. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/add-tools))
- [Agentic Workflow Graphs](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-workflow-graphs.md) — Models task sequences as directed graphs to support conditional routing and parallel execution of autonomous agents.
- [AI Agent Orchestration Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/ai-agent-orchestration-frameworks.md) — Implements runtimes that manage agent loops and interact with underlying inference services via direct configuration. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/your-first-agent))
- [LLM Tooling Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/artificial-intelligence-tooling/language-model-integrations/llm-tooling-integrations.md) — Provides a dedicated integration layer for binding external functions and APIs as executable tools for LLMs.
- [Conversation History Management](https://awesome-repositories.com/f/artificial-intelligence-ml/conversation-history-management.md) — Stores and manages logs of agent-user interactions to maintain multi-turn context. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/memory))
- [Conversation State Management](https://awesome-repositories.com/f/artificial-intelligence-ml/conversation-state-management.md) — Implements conversational state management to persist chat history and session memory across interactions.
- [Conversational Session Management](https://awesome-repositories.com/f/artificial-intelligence-ml/conversational-session-management.md) — Maintains conversation history across multi-turn dialogues by mapping unique identifiers to session stores.
- [Conversational State Managers](https://awesome-repositories.com/f/artificial-intelligence-ml/conversational-state-managers.md) — Provides a state persistence layer that tracks session history and conversation threads across various stores.
- [Human Approval Gates](https://awesome-repositories.com/f/artificial-intelligence-ml/human-approval-gates.md) — Provides human-in-the-loop approval gates to authorize sensitive tool invocations before the model processes the results. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/tools/index))
- [Persistent Chat Histories](https://awesome-repositories.com/f/artificial-intelligence-ml/language-model-orchestration/ai-memory-systems/persistent-chat-histories.md) — Persists conversation records using NoSQL or key-value stores to maintain long-term history across sessions. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/index))
- [Conversation Threads](https://awesome-repositories.com/f/artificial-intelligence-ml/language-model-orchestration/conversation-management/conversation-threads.md) — Provides persistent containers for organizing user-agent message history as conversation threads. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-semantic-kernel/index))
- [LLM Provider Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/llm-provider-integrations.md) — Provides connectivity and configuration for various inference services including cloud APIs and local open-source models. ([source](https://cdn.jsdelivr.net/gh/microsoft/agent-framework@main/README.md))
- [Retrieval-Augmented Generation Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/retrieval-augmented-generation-frameworks.md) — Includes a framework for executing retrieval-augmented generation pipelines to ground model responses in external data. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/index))
- [Stateful Run Executions](https://awesome-repositories.com/f/artificial-intelligence-ml/stateful-agent-orchestration/global-run-local-state/stateful-run-executions.md) — Implements the runtime execution of agents while maintaining conversation state across sequential steps. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/run-agent/))
- [Human-in-the-Loop Approvals](https://awesome-repositories.com/f/artificial-intelligence-ml/step-based-schedulers/step-execution-engines/execution-step-controllers/human-in-the-loop-approvals.md) — Implements mechanisms to pause autonomous execution and require human approval before tool invocation.
- [Sequential Step Orchestrators](https://awesome-repositories.com/f/artificial-intelligence-ml/step-based-schedulers/step-execution-engines/execution-step-controllers/sequential-step-orchestrators.md) — Links multiple processing steps in a sequential chain where data flows from one executor to the next. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/workflows))
- [Agent Connectivity Interfaces](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-connectivity-interfaces.md) — Implements protocols for resolving agent capabilities via metadata and establishing communication across remote endpoints. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/a2a))
- [Agent Definitions](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-definitions.md) — Allows defining specialized agent behaviors and roles by extending core base abstractions. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/index))
- [Agent Function Libraries](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-function-libraries.md) — Provides agents with libraries of specialized external functions to perform tasks and access real-time data. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/index))
- [Session Personalization](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-session-management/session-personalization.md) — Injects personalized user information into agent runs via context providers to tailor behavior. ([source](https://learn.microsoft.com/en-us/agent-framework/get-started/memory))
- [Agent Skill Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-skill-frameworks.md) — Provides a flexible system for loading agent capabilities from files, inline code, or classes. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/skills))
- [Agent State Persistence](https://awesome-repositories.com/f/artificial-intelligence-ml/agent-state-persistence.md) — Saves and restores the execution state of autonomous agents to ensure continuity after failures or restarts. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/durable-extension))
- [Agent Capability Extensions](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-capabilities-skills-tooling/agent-capability-extensions.md) — Provides reusable modules that package domain expertise to expand the functional capabilities of agents. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/skills))
- [Agent System Prompts](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-capabilities-skills-tooling/agent-system-prompts.md) — Allows customization of system prompts to control how agent skills and tools are presented to the model. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/skills))
- [Agent Configurations](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents/agent-configurations.md) — Uses typed dictionaries to define and manage inference parameters and provider settings for agents. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-semantic-kernel/index))
- [Conversational Agent Construction](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/agent-orchestration-multi-agent/autonomous-agents/ai-agent-builders/agent-construction-frameworks/conversational-agent-construction.md) — Supports the construction of agents for multi-turn conversations with managed history and structured outputs. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/index))
- [Agent Tooling Definitions](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/integration-deployment/agent-frameworks/configuration-and-specifications/agent-tooling-definitions.md) — Provides mechanisms to define tool signatures that allow agents to request execution on remote systems. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/function-tools/))
- [Domain Expertise Packages](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/integration-deployment/agentic-domains/domain-expertise-packages.md) — Provides modular domain expertise packages that agents load to perform specialized professional tasks. ([source](https://cdn.jsdelivr.net/gh/microsoft/agent-framework@main/README.md))
- [OpenAI-Compatible APIs](https://awesome-repositories.com/f/artificial-intelligence-ml/agentic-systems-frameworks/model-integration-serving/model-integration-interfaces/ai-integration-apis/openai-compatible-apis.md) — Exposes a backend API that follows the OpenAI specification for compatibility with standard AI SDKs. ([source](https://learn.microsoft.com/en-us/agent-framework/devui/index))
- [Agent Message Proxies](https://awesome-repositories.com/f/artificial-intelligence-ml/ai-agent-connectors/agent-message-proxies.md) — Enables connection and invocation of agents hosted on remote services through standardized message proxying. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/index))
- [Workflow Validation](https://awesome-repositories.com/f/artificial-intelligence-ml/ai-agent-workflow-definition/workflow-validation.md) — Verifies graph connectivity and type compatibility to ensure the execution graph is logically sound. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/workflows))
- [Runtime Context Injections](https://awesome-repositories.com/f/artificial-intelligence-ml/context-injection/runtime-context-injections.md) — Passes session-specific data to tool functions without exposing internal parameters to the model's schema. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/function-tools/))
- [Context Window Optimizations](https://awesome-repositories.com/f/artificial-intelligence-ml/context-window-optimizations.md) — Optimizes token usage by implementing high-level summaries and on-demand loading of detailed instructions. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/skills))
- [Session History Retrieval](https://awesome-repositories.com/f/artificial-intelligence-ml/conversational-session-management/multi-session-context-synthesizers/session-history-retrieval.md) — Reconstructs conversation state and retrieves historical data using service identifiers for context restoration. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/agents/multi-turn-conversation/))
- [Custom Provider Implementations](https://awesome-repositories.com/f/artificial-intelligence-ml/custom-provider-implementations.md) — Provides a base class to implement custom agent types for proprietary or unsupported inference services. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/providers/index))
- [External Service Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/external-service-integrations.md) — Connects workflows to external APIs and integrates human-in-the-loop interaction patterns. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/workflows/overview/))
- [External System Integrations](https://awesome-repositories.com/f/artificial-intelligence-ml/external-system-integrations.md) — Integrates agent logic with durable extensions, communication protocols, and developer interfaces. ([source](https://learn.microsoft.com/en-us/agent-framework/))
- [External Tool Execution](https://awesome-repositories.com/f/artificial-intelligence-ml/external-tool-execution.md) — Executes function tools, code interpreters, and file search capabilities provided by the inference service. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/providers/index))
- [Function-to-Tool Converters](https://awesome-repositories.com/f/artificial-intelligence-ml/function-to-tool-converters.md) — Converts standard code methods into executable agent tools with automatic schema generation. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/function-tools/))
- [Hierarchical Agent Orchestration](https://awesome-repositories.com/f/artificial-intelligence-ml/hierarchical-agent-orchestration.md) — Supports hierarchical agent structures where one agent can be wrapped as a tool for another. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/tools/index))
- [Human-in-the-Loop Workflows](https://awesome-repositories.com/f/artificial-intelligence-ml/human-in-the-loop-workflows.md) — Incorporates human-in-the-loop workflows with approval gates to authorize tool use or review agent actions.
- [Inference Configuration Parameters](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-inference-serving/model-integration-pipelines/model-inference/inference-configuration-parameters.md) — Sets provider-specific inference parameters, such as output tokens, using typed configuration dictionaries. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-semantic-kernel))
- [Model Provider Abstractions](https://awesome-repositories.com/f/artificial-intelligence-ml/model-provider-abstractions.md) — Decouples agent logic from specific LLM providers using a base class abstraction layer.
- [Schema Enforcement Layers](https://awesome-repositories.com/f/artificial-intelligence-ml/model-response-parsers/schema-enforcement-layers.md) — Constrains model outputs to adhere to specific JSON schemas during the generation process. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/structured-output/))
- [Multi-Agent Orchestrators](https://awesome-repositories.com/f/artificial-intelligence-ml/multi-agent-orchestrators.md) — Coordinates teams of specialized agents to automate complex business processes via sequential or concurrent patterns. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/workflows/overview/))
- [Skill Availability Controls](https://awesome-repositories.com/f/artificial-intelligence-ml/skill-availability-controls.md) — Allows adding or restricting available tools during an active session based on the current agent state. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/function-tools/))
- [Structured Output Converters](https://awesome-repositories.com/f/artificial-intelligence-ml/structured-output-converters.md) — Uses a transformation layer to convert plain text model responses into validated JSON schemas.
- [Tool Calling](https://awesome-repositories.com/f/artificial-intelligence-ml/tool-calling.md) — Enables agents to iteratively execute external functions until a task is complete, including loop safety mechanisms. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen/index))
- [Tool Schema Definitions](https://awesome-repositories.com/f/artificial-intelligence-ml/tool-schema-definitions.md) — Implements standardized formats for function names and parameter types to guide model tool selection. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/function-tools/))
- [Streaming Workflow Execution](https://awesome-repositories.com/f/artificial-intelligence-ml/workflow-orchestration/streaming-workflow-execution.md) — Streams final answers and intermediate progress updates back to the caller during workflow execution. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/workflows))

### Software Engineering & Architecture

- [Multi-Agent Coordination Frameworks](https://awesome-repositories.com/f/software-engineering-architecture/multi-agent-coordination-frameworks.md) — Coordinates the execution of multiple agents using directed graphs to manage data flow and task orchestration. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/workflows))
- [Control Flow Logic Models](https://awesome-repositories.com/f/software-engineering-architecture/control-flow-logic-models.md) — Defines execution paths via directed graphs or functional logic to implement conditional routing. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/workflows/overview/))
- [Graph-Based Workflow Models](https://awesome-repositories.com/f/software-engineering-architecture/graph-based-workflow-models.md) — Uses graph architectures to model complex task sequences supporting parallel processing and conditional routing. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/index))
- [Workflow Logic Engines](https://awesome-repositories.com/f/software-engineering-architecture/workflow-logic-engines.md) — Implements the underlying execution logic and graph edges required to coordinate complex agent task sequences. ([source](https://learn.microsoft.com/en-us/agent-framework/))
- [Typed Message Processing](https://awesome-repositories.com/f/software-engineering-architecture/workflow-triggers/inter-workflow-messaging/typed-message-processing.md) — Executes custom logic by receiving and processing typed messages within dedicated workflow processing units. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/executors))
- [Action Dispatch Middleware](https://awesome-repositories.com/f/software-engineering-architecture/action-dispatch-middleware.md) — Applies middleware to agent operations to implement telemetry, safety mitigations, and business logic. ([source](https://learn.microsoft.com/en-us/agent-framework/overview/))
- [Background Task Management](https://awesome-repositories.com/f/software-engineering-architecture/background-task-management.md) — Executes processes in the background and provides continuation tokens to poll for results or subscribe to updates. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/a2a))
- [Dependency Injection Containers](https://awesome-repositories.com/f/software-engineering-architecture/dependency-injection-containers.md) — Ships a central service registry to manage shared resources, tools, and agent configurations via dependency injection.
- [Dependency Injection](https://awesome-repositories.com/f/software-engineering-architecture/integration-extensibility/dependency-injection.md) — Injects service providers into skill resources to allow agents access to external data and business logic. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/skills))
- [Operation Interceptors](https://awesome-repositories.com/f/software-engineering-architecture/request-interception-middleware/operation-interceptors.md) — Intercepts internal agent operations to implement input validation, content filtering, logging, and caching. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen/index))
- [Request Middleware](https://awesome-repositories.com/f/software-engineering-architecture/request-middleware.md) — Implements a request middleware pipeline to handle cross-cutting concerns like security filtering and rate limiting. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen))

### Data & Databases

- [Vector Database Abstractions](https://awesome-repositories.com/f/data-databases/database-abstraction-layers/vector-database-abstractions.md) — Provides a unified abstraction layer to connect agents to multiple vector database backends for efficient data retrieval. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/index))
- [Message Routing](https://awesome-repositories.com/f/data-databases/message-brokers/message-routing.md) — Directs workflow message flow using conditional branching and parallel fan-out patterns. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/edges))
- [State Checkpointing](https://awesome-repositories.com/f/data-databases/state-checkpointing.md) — Records the execution state of graphs to avoid repeating completed steps after a system failure. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/durable-extension))
- [Session State Management](https://awesome-repositories.com/f/data-databases/session-state-management.md) — Persists conversation history and session state using external stores such as Redis. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen/index))
- [Text-to-JSON Converters](https://awesome-repositories.com/f/data-databases/text-to-json-converters.md) — Transforms plain text responses into structured JSON using a decorator pattern for models without native support. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/structured-output/))

### Networking & Communication

- [In-Process Message Routing](https://awesome-repositories.com/f/networking-communication/in-process-message-routing.md) — Implements internal message routing between processing units to coordinate data flow within the agent framework. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/executors))
- [Middleware-Based Request Pipelines](https://awesome-repositories.com/f/networking-communication/communication-protocols-architectures/request-processing-architectures/request-processing/middleware-based-request-pipelines.md) — Provides a modular chain of handlers to intercept agent operations for logging, safety filtering, and telemetry.
- [Agent Runtime Exposure](https://awesome-repositories.com/f/networking-communication/http-clients/http-servers/agent-runtime-exposure.md) — Wraps internal agents in a standardized server-side protocol for discovery and invocation by external clients. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/a2a))
- [Agent Telemetry Streams](https://awesome-repositories.com/f/networking-communication/real-time-telemetry-streams/agent-telemetry-streams.md) — Implements real-time telemetry streams to track agent logs, request metadata, and token consumption. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen))

### Development Tools & Productivity

- [Human-in-the-loop Interfaces](https://awesome-repositories.com/f/development-tools-productivity/human-in-the-loop-interfaces.md) — Suspends execution to wait for external events or human approval without consuming active compute resources. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/durable-extension))
- [Tool Function Registrations](https://awesome-repositories.com/f/development-tools-productivity/local-function-execution/agent-integrated-functions/tool-function-registrations.md) — Binds functions directly to an agent during creation or execution to extend capabilities without specialized attributes. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-semantic-kernel/index))
- [Session State Serializers](https://awesome-repositories.com/f/development-tools-productivity/session-capturers/session-state-serializers.md) — Serializes conversation data and agent states to allow restoration from stored snapshots. ([source](https://learn.microsoft.com/en-us/agent-framework/agents/conversations/index))

### Security & Cryptography

- [Agent Endpoint Access Control](https://awesome-repositories.com/f/security-cryptography/agent-endpoint-access-control.md) — Intercepts requests to agent endpoints to verify identities and enforce access control policies. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/a2a))
- [Session Identifiers](https://awesome-repositories.com/f/security-cryptography/identity-access-management/session-management/session-identifiers.md) — Uses unique session identifiers to reconnect agents to existing conversations and reload state. ([source](https://learn.microsoft.com/en-us/agent-framework/tutorials/agents/multi-turn-conversation/))
- [Agentic Session Persistence](https://awesome-repositories.com/f/security-cryptography/identity-access-management/session-management/stateful-session-persistence/agentic-session-persistence.md) — Tracks task progress and session state in external stores to allow agents to resume work across sessions. ([source](https://learn.microsoft.com/en-us/agent-framework/overview/agent-framework-overview/))

### System Administration & Monitoring

- [Agent Execution Tracing](https://awesome-repositories.com/f/system-administration-monitoring/agent-execution-tracing.md) — Ships a dashboard to visualize agent reasoning trajectories and tool usage for performance analysis. ([source](https://learn.microsoft.com/en-us/agent-framework/integrations/durable-extension))
- [Agent Observability](https://awesome-repositories.com/f/system-administration-monitoring/agent-observability.md) — Includes telemetry and tracing capabilities to monitor agent performance and debug complex multi-step workflows.
- [Agent Performance Monitoring](https://awesome-repositories.com/f/system-administration-monitoring/agent-performance-monitoring.md) — Tracks operational metrics and distributed tracing data to observe and optimize agent behavior. ([source](https://cdn.jsdelivr.net/gh/microsoft/agent-framework@main/README.md))
- [Event Monitoring Systems](https://awesome-repositories.com/f/system-administration-monitoring/event-monitoring-systems.md) — Tracks workflow progress through a system of lifecycle events and real-time state broadcasting. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/workflows/overview/))
- [AI and Agent Observability](https://awesome-repositories.com/f/system-administration-monitoring/monitoring-and-observability/ai-agent-observability.md) — Ships an observability suite using OpenTelemetry to trace execution flows and monitor tool calls.
- [GenAI Execution Monitoring](https://awesome-repositories.com/f/system-administration-monitoring/monitoring-and-observability/workflow-performance-diagnostics/genai-execution-monitoring.md) — Provides specialized observability for tracking the internal execution flow and runtime behavior of generative AI components. ([source](https://learn.microsoft.com/en-us/agent-framework/workflows/index))
- [Observability Instrumentation](https://awesome-repositories.com/f/system-administration-monitoring/observability-instrumentation.md) — Instruments AI workflows via environment variables and configuration to enable distributed tracing and telemetry. ([source](https://learn.microsoft.com/en-us/agent-framework/migration-guide/from-autogen/index))
- [Agent State Tracking](https://awesome-repositories.com/f/system-administration-monitoring/system-activity-monitoring/session-activity-monitors/agent-state-tracking.md) — Tracks session-based memory and internal state transitions for long-running tasks and human interactions. ([source](https://learn.microsoft.com/en-us/agent-framework/user-guide/overview))

### Part of an Awesome List

- [Agent Frameworks](https://awesome-repositories.com/f/awesome-lists/ai/agent-frameworks.md) — Framework for building and deploying multi-agent workflows.
- [AI Agent Frameworks](https://awesome-repositories.com/f/awesome-lists/ai/ai-agent-frameworks.md) — Framework for orchestrating multi-agent workflows.