Agent Development Kit (ADK) MOC
This map organizes the Agent Development Kit (ADK) ecosystem, a code-first toolkit for building and deploying AI agents.
🏗️ Core Architecture
- agent-development-kit - The primary entry point and architectural blueprint.
- adk-callbacks-and-lifecycle - The Runner model, events, and execution hooks.
- adk-go-implementation - Reference implementation in Go.
🛠️ Services & Primitives
- adk-session-service - Context, history, and short-term state.
- adk-artifact-service - Versioned binary data management.
- adk-long-term-memory - Cross-session recall and knowledge storage.
🚦 Orchestration & Coordination
- adk-multi-agent-orchestration - Hierarchical teams and handoff mechanisms.
- workflow-agents - Deterministic control flow (Sequential, Parallel, Loop).
- adk-advanced-capabilities - Planning (
ThinkingConfig), code execution, and streaming/live audio (Runner.run_live,LiveRequestQueue, native-audio models).
📈 Evaluation & Maintenance
- adk-evaluation-framework - Benchmarking and multi-turn dataset creation.
- agent-observability - Traces, spans, and lifecycle hooks for production monitoring.
- lit-adk-documentation - Core literature source for ADK.
References
Capabilities & Features
- telephony-agents - Autonomous telephony via AgentPhone and streaming voice interactions.
- lit-adk-telephony - Technical specifications for AgentPhone MCP and Gemini Live streaming.