Ck:Google Adk Python

Google ADK Python Skill

Expert guide for Google's Agent Development Kit (ADK) Python — open-source, code-first toolkit for building, evaluating, and deploying AI agents. Optimized for Gemini, model-agnostic by design.

When to Activate

• Build single or multi-agent systems with tool integration
• Implement A2A protocol for remote agent communication
• Integrate MCP servers as agent tools
• Use workflow agents (sequential, parallel, loop) for pipelines
• Manage sessions, state, memory, and artifacts
• Add callbacks, plugins, or observability hooks
• Deploy to Cloud Run, Vertex AI Agent Engine, or GKE
• Evaluate agents with adk eval framework

Agent Structure Convention (Required)

Copy code
my_agent/
├── __init__.py   # MUST: from . import agent
└── agent.py      # MUST: root_agent = Agent(...) OR app = App(...)

Quick Start

bashCopy code
pip install google-adk          # stable (weekly releases)
uv sync --all-extras            # dev setup (uv required, Python 3.10+, 3.11+ recommended)

pythonCopy code
from google.adk import Agent

root_agent = Agent(
    name="assistant",
    model="gemini-2.5-flash",
    instruction="You are a helpful assistant.",
    description="General assistant agent.",
    tools=[get_weather],
)

App Pattern (Production)

pythonCopy code
from google.adk import Agent
from google.adk.apps import App
from google.adk.apps.app import EventsCompactionConfig
from google.adk.plugins.save_files_as_artifacts_plugin import SaveFilesAsArtifactsPlugin

app = App(
    name="my_app",
    root_agent=Agent(name="my_agent", model="gemini-2.5-flash", ...),
    plugins=[SaveFilesAsArtifactsPlugin()],
    events_compaction_config=EventsCompactionConfig(compaction_interval=2),
)

Use App when needing plugins, event compaction, or custom lifecycle management.

CLI Tools

Agent Types

Key APIs

Model Support

Latest: gemini-2.5-flash (default), gemini-2.5-pro, gemini-2.0-flash (sunsets Mar 2026) Preview: gemini-3-flash-preview, gemini-3-pro-preview Also: OpenAI Codex, Ollama, LiteLLM, vLLM, Model Garden

Best Practices

1. Code-first — define agents in Python for version control and testing
2. Agent convention — always use root_agent or app variable in agent.py
3. Modular agents — specialize per domain, compose via sub_agents
4. Workflow selection — workflow agents for predictable, LlmAgent for dynamic
5. State — ToolContext.state for ephemeral, MemoryService for long-term
6. Safety — callbacks for guardrails, tool confirmation for sensitive ops
7. Evaluate — test with adk eval + evalset JSON before deployment

References

Detailed guides (load as needed):

• references/agent-types-and-architecture.md — Agent types, workflows, custom agents
• references/tools-and-mcp-integration.md — Custom tools, MCP, tool filtering
• references/multi-agent-and-a2a-protocol.md — Sub-agents, A2A, coordinator patterns
• references/sessions-state-memory-artifacts.md — State, artifacts, sessions, memory
• references/callbacks-plugins-observability.md — Lifecycle hooks, plugins, tracing
• references/evaluation-testing-cli.md — adk eval, CLI, evalset format
• references/deployment-cloud-run-vertex-gke.md — Cloud Run, Vertex AI, GKE

External Resources

• GitHub: https://github.com/google/adk-python
• Docs: https://google.github.io/adk-docs/
• Samples: https://github.com/google/adk-python/tree/main/contributing/samples
• llms.txt: https://raw.githubusercontent.com/google/adk-python/refs/heads/main/llms.txt