google/adk-samples
Adk Samples
This project provides a collection of reference implementations, architectural patterns, and SDK samples for building autonomous agents using large language models. It serves as a multi-language framework for implementing and deploying specialized AI agents across diverse programming environments.
The system centers on an orchestration framework that combines deterministic code with adaptive reasoning through structured graph workflows. It utilizes schema-driven integration to connect agents with third-party applications and diverse AI models.
The development lifecycle is supported by toolkits for measuring agent reliability and performance against quality benchmarks. A command line interface manages the transition of these agents from initial prototyping and testing to production deployment.
Features
- AI Agent Development - Provides the foundational framework, reference architectures, and tools for developing specialized autonomous AI agents.
- AI Agent Architectures - Organizes agent logic into modular architectures that decouple reasoning from service implementations.
- Agentic Workflow Graphs - Uses structured graph architectures to combine deterministic code with adaptive reasoning for complex task management.
- Agentic Workflow Orchestration - Orchestrates complex task execution paths by combining adaptive reasoning with structured graph workflows.
- AI Agent Tool Integrations - Connects AI agents to external software, databases, and APIs to extend their data access and functionality.
- AI Agent Orchestration Frameworks - Implements a runtime for managing autonomous agent loops and tool discovery within durable pipelines.
- Autonomous Agent Patterns - Provides architectural patterns and reference implementations for designing specialized autonomous AI agents.
- Multi-Language Agent SDKs - Provides a set of programming samples and SDKs for implementing AI agents across diverse coding environments.
- Tool-Calling Schemas - Maps third-party API capabilities to structured schemas that enable AI models to call external functions.
- Tool-Use Integrations - Enables agents to expand their functional capabilities by integrating third-party applications and diverse AI models.
- LLM Reasoning Workflows - Implements graph-based orchestrators designed for sequences of LLM operations and adaptive reasoning patterns.
- Agent Deployment - Provides a CLI for the creation, testing, and provisioning of autonomous agent instances.
- Agent Deployment Tools - Implements utilities for transitioning AI agents from initial local prototypes to production environments.
- Agent Evaluation Tools - Provides a specialized testing suite to assess reasoning and tool usage in autonomous agent workflows.
- Evaluation Metrics - Provides methods for measuring agent reliability by comparing execution outputs against quality benchmarks.
- AI Observability and Evaluation - Offers tools for evaluating agent reliability and effectiveness before production deployment.
- Deployment Command Line Interfaces - Ships a command line interface to manage the transition of agents from prototyping to production deployment.
- Cross-Language Implementations - Applies a consistent set of design patterns across multiple programming languages to ensure cross-platform compatibility.
