autogen-vertexai-memory 0.1.3

VertexAI Memory integration for Autogen agents

autogen-vertexai-memory

VertexAI Memory integration for Autogen agents. This package provides a Memory implementation that enables Autogen agents to store and retrieve memories using Google Cloud's VertexAI Memory service.

Features

• 🧠 Persistent Memory: Store agent memories in VertexAI for long-term retention
• 🔍 Semantic Search: Retrieve relevant memories using similarity search
• 🔄 Context Management: Automatically update chat context with relevant memories
• ⚡ Async Support: Built with async/await for efficient operations
• 🎯 User Scoping: Isolate memories per user for multi-tenant applications

Installation

pip install autogen-vertexai-memory

Or with poetry:

poetry add autogen-vertexai-memory

Prerequisites

1. Google Cloud Project: You need a GCP project with VertexAI API enabled
2. Authentication: Set up authentication using one of these methods:
   • Application Default Credentials (recommended for local development)
   • Service Account Key
   • Workload Identity (for GKE)

# Set up Application Default Credentials
gcloud auth application-default login

3. VertexAI Memory Resource: Create a memory resource in your GCP project

Quick Start

from autogen_vertexai_memory import VertexaiMemory
from autogen_core.memory import MemoryContent, MemoryMimeType

# Initialize the memory
memory = VertexaiMemory(
    api_resource_name="projects/YOUR_PROJECT/locations/us-central1/memories/YOUR_MEMORY",
    project_id="YOUR_PROJECT_ID",
    location="us-central1"
)

# Add a memory
await memory.add(
    content=MemoryContent(
        content="The user prefers Python over JavaScript",
        mime_type=MemoryMimeType.TEXT
    ),
    user_id="user123"
)

# Query memories
results = await memory.query(
    query="What programming language does the user prefer?",
    user_id="user123"
)

for result in results.results:
    print(result.content)

API Reference

VertexaiMemory

Constructor

VertexaiMemory(
    api_resource_name: str,
    project_id: str,
    location: str,
    client: Client | None = None
)

Parameters:

• api_resource_name: Full resource name of the VertexAI memory
• project_id: Google Cloud project ID
• location: GCP region (e.g., "us-central1")
• client: Optional pre-configured VertexAI client

Methods

add(content, user_id, cancellation_token=None)

Add a new memory.

await memory.add(
    content=MemoryContent(content="Important fact", mime_type=MemoryMimeType.TEXT),
    user_id="user123"
)

query(query="", user_id="", cancellation_token=None, **kwargs)

Query memories. If query is empty, returns all memories for the user.

# Search with query
results = await memory.query(query="user preferences", user_id="user123")

# Get all memories
all_memories = await memory.query(user_id="user123")

update_context(model_context)

Update chat context with relevant memories.

from autogen_core.model_context import ChatCompletionContext

context = ChatCompletionContext()
await memory.update_context(context)

clear()

Clear all memories (deletes the resource).

await memory.clear()

close()

Cleanup resources.

await memory.close()

Advanced Usage

Custom Client Configuration

from vertexai import Client

# Create custom client with specific settings
client = Client(
    project="my-project",
    location="us-central1",
    # Additional client configurations...
)

# Use with VertexaiMemory
memory = VertexaiMemory(
    api_resource_name="...",
    project_id="my-project",
    location="us-central1",
    client=client
)

Batch Operations

# Add multiple memories
memories = [
    ("User likes cats", "user123"),
    ("User prefers dark mode", "user123"),
    ("User is from New York", "user123")
]

for content, user_id in memories:
    await memory.add(
        content=MemoryContent(content=content, mime_type=MemoryMimeType.TEXT),
        user_id=user_id
    )

Error Handling

from google.api_core.exceptions import GoogleAPIError

try:
    await memory.add(content=..., user_id="user123")
except GoogleAPIError as e:
    print(f"Error adding memory: {e}")

Development

Setup Development Environment

# Clone the repository
git clone https://github.com/yourusername/autogen-vertexai-memory.git
cd autogen-vertexai-memory

# Install dependencies
poetry install

# Run tests
poetry run pytest

Running Tests

# Run all tests
poetry run pytest tests/ -v

# Run with coverage
poetry run pytest tests/ --cov=autogen_vertexai_memory --cov-report=html

# Run specific test
poetry run pytest tests/test_vertexai_memory.py::TestAdd -v

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Built for use with Autogen
• Powered by Google Cloud VertexAI

Support

• 📫 Report issues on GitHub Issues
• 💬 Discuss on GitHub Discussions
• 📖 Read the documentation

Changelog

See CHANGELOG.md for a list of changes.

---

Made with ❤️ for the Autogen community