google/adk-go

Features

• Agent Orchestration Engines - Manages the lifecycle, sequencing, and execution flow of autonomous agent tasks across cloud environments.
• Agent Tool Integrations - Provides mechanisms for connecting autonomous agents to external software tools, APIs, and services.
• Agentic Workflow Automation - Coordinates complex tasks using a blend of deterministic code and adaptive AI reasoning.
• Agentic Workflow Graphs - Directs agent execution paths using directed graphs that combine deterministic logic with adaptive reasoning.
• LLM Tooling Integrations - Offers connectors and interfaces that allow language models to access external data and execute software tools.
• Multi-Agent Frameworks - Provides a framework for coordinating multiple autonomous agents into collaborative teams using structured delegation.
• Multi-Agent Orchestration Systems - Provides a platform for coordinating multiple autonomous agents to execute complex, collaborative workflows.
• Multi-Agent System Composition - Implements structures for nesting and coordinating specialized autonomous agents into modular systems.
• Agent Evaluation Tools - Provides structured schemas and tools to measure the reliability and accuracy of agent outputs.
• External Tool Integration - Enables agents to interact with external APIs and human-in-the-loop processes to extend functional capabilities.
• Multi-Agent Output Evaluation - Measures agent reliability by comparing model responses against predefined schemas and validation tools.
• Tool Integration Interfaces - Provides standardized interfaces to expose functions as executable tools for language models.
• Cloud Agent Deployers - Automates containerization and registration of agent services for scalable cloud hosting.
• Serverless Runtimes - Provides a runtime model that decouples agent logic from host environments for rapid serverless scaling.

Star history