Skip to content

Latest commit

 

History

History

google_adk

README.md

Google ADK + Nexus Integration

This example shows how to integrate Nexus filesystem operations with Google ADK agents.

Quick Start

1. Install Dependencies

pip install google-adk nexus-ai-fs

2. Get Google API Key

Get your Gemini API key at: https://makersuite.google.com/app/apikey

3. Run the Example

export GOOGLE_API_KEY="your-key-here"
python examples/google_adk/basic_adk_agent.py

Or use the helper script:

export GOOGLE_API_KEY="your-key-here"
bash RUN_GOOGLE_ADK.sh

What It Does

The agent can:

  • Search file content with grep_files()
  • Find files by name with glob_files()
  • Read file content with read_file()
  • Write reports with write_file()

Example task: Find all Python files with async/await, analyze them, and write a summary report.

How It Works

Google ADK requires:

  1. Agent - LlmAgent with tools and instructions
  2. Session Service - Manages conversation state (InMemorySessionService)
  3. Runner - Executes agent with session management
  4. Session Creation - Must create session before running
from google.adk.agents import LlmAgent
from google.adk import Runner
from google.adk.sessions import InMemorySessionService, Session

# 1. Create agent with Nexus tools
agent = LlmAgent(
    name="file_agent",
    model="gemini-2.5-flash",
    instruction="You are a filesystem assistant",
    tools=[grep_files, read_file, write_file]
)

# 2. Create session service
session_service = InMemorySessionService()

# 3. Create runner
runner = Runner(
    app_name="my-app",
    agent=agent,
    session_service=session_service
)

# 4. Create session
session = Session(
    id="session-123",
    user_id="user-123",
    app_name="my-app"
)
session_service.create(session)

# 5. Run with formatted message
message = types.Content(
    role="user",
    parts=[types.Part(text="Find Python files")]
)

for event in runner.run(
    user_id="user-123",
    session_id="session-123",
    new_message=message
):
    # Process events...
    pass

Comparison with LangGraph

Both frameworks integrate with Nexus the same way - wrap Nexus operations as Python functions and pass as tools.

Key Differences:

Aspect LangGraph Google ADK
Setup StateGraph + nodes Agent + Runner + SessionService
API Style Graph-based Service-based
Session Management Optional Required
Best For Complex workflows, multi-LLM Production deployments, Google Cloud

Files

  • basic_adk_agent.py - Main example (✅ working)
  • multi_agent_demo.py - Multi-agent coordination (conceptual)
  • test_installation.py - Verify setup
  • comparison_with_langgraph.md - Detailed comparison

Troubleshooting

Import Errors

Google ADK uses namespace packages which can conflict in virtual environments. If you get import errors:

# Option 1: Use global Python
/opt/anaconda3/bin/python examples/google_adk/basic_adk_agent.py

# Option 2: Fresh conda environment
conda create -n adk python=3.12
conda activate adk
pip install google-adk nexus-ai-fs

API Key Errors

Make sure GOOGLE_API_KEY is set:

export GOOGLE_API_KEY="your-key"
python examples/google_adk/basic_adk_agent.py

Next Steps

  • Try the working example: python examples/google_adk/basic_adk_agent.py
  • Compare with LangGraph: python examples/langgraph/langgraph_react_demo.py
  • Check Claude Agent SDK (simpler): python examples/claude_agent_sdk/claude_agent_react_demo.py

Questions?