ADK Developer Guides
This directory contains specific developer guides for the ADK Python implementation. For the official ADK documentation, visit
.
Index
Agents
Creating Agents with Configurations
- Building and wiring multi-agent graphs from external YAML configuration files.
- Guide on using LlmAgent in single-turn mode.
- Guide on using LlmAgent in task mode.
- Guide on using ManagedAgent with server-side tools.
- Guide on using RemoteA2aAgent in task mode.
Apps
- The top-level container binding a root agent to app-wide plugins and configuration.
Artifacts
- Storing binary payloads outside the conversation history, with versioning and user-scoped filenames.
Auth
AuthConfig and authenticated tools
- Declaring the credentials a tool needs, and the pause-for-consent handshake.
Code Executors
- Executing model-generated code safely across local, container, GKE, and managed sandbox backends.
Events
- Understanding Event and NodeInfo in workflows.
- How to use RequestInput for human-in-the-loop interactions.
Flows
- Inspecting or blocking content on a live bidirectional session.
Integrations
- Screening user input and model output with Google Cloud Model Armor.
Labs
- Runs a Google Antigravity SDK agent as an ADK agent node.
Live
- Streaming content, realtime audio, and stream control signals to live agents.
Memory
- Storing finished sessions and recalling them from later conversations.
Models
- The model interface, how a model name resolves to an implementation, and how to plug in your own.
Planners
- Guiding model execution with structured planning instructions, thinking configurations, and Plan-Re-Act thought tagging.
Plugins
- Self-healing, concurrent-safe error recovery for model failures.
- Self-healing, concurrent-safe error recovery for tool failures.
Runners
- Managing session lifecycles, state resolution, and streaming agent execution events.
- Real-time bidirectional audio/text streaming and non-blocking background tool execution with Gemini Multimodal Live API.
Sessions
Session and BaseSessionService
- The session lifecycle, state scoping, and choosing a session service.
- Session state and the app:, user:, and temp: prefixes that decide what is shared and what is stored.
Tools
- Exposing workflows and deterministic nodes as agent tools with isolated runtime branching and resume support.
- Expose an ADK agent as an MCP server so any MCP host can drive it as a single tool (the MCP counterpart of to_a2a).
Workflows
- Graph-based orchestration of complex, multi-step agent interactions.
- Understanding nodes, edges, and graph structures in workflows.
- Wrapping Python functions and generators as workflow nodes.
- Synchronizing parallel execution paths in workflows.
- Configuring retry policies for resilient workflow nodes.
- Processing lists of items concurrently in workflows.
- Scheduling and executing nodes dynamically at runtime.