chainlit-sqlalchemy

SQLAlchemy data layer for Chainlit.

SQLAlchemy Data Layer

This data layer supports PostgreSQL and other SQL databases via SQLAlchemy.

Key features:

• Supports storage clients for attachments (currently: Azure, Azure Blobs, GCS, S3)
• User/thread/step/element/feedback storage in SQL
• Async operations

Setup Example (PostgreSQL + Azure Blob)

1. Load schema.sql into your database.
2. Install required dependencies:

# For PostgreSQL
pip install chainlit-sqlalchemy[postgres]

# For SQLite
pip install chainlit-sqlalchemy[sqlite]

# With cloud storage
pip install chainlit-sqlalchemy[postgres,gcs]    # PostgreSQL + Google Cloud Storage
pip install chainlit-sqlalchemy[sqlite,azure-blob]  # SQLite + Azure Blob

2. Configure in your Chainlit app:

import chainlit as cl
from chainlit.data.sql_alchemy import SQLAlchemyDataLayer
from chainlit.data.storage_clients import AzureBlobStorageClient

@cl.data_layer
def get_data_layer():
    storage_client = AzureBlobStorageClient(
        container_name="<your_container>",
        storage_account="<your_account>",
        storage_key="<your_key>"
    )

    return SQLAlchemyDataLayer(
        conninfo="postgresql+asyncpg://user:password@host/dbname",
        storage_provider=storage_client
    )

Dependencies

• Core: SQLAlchemy
• Database Drivers (optional):
  • PostgreSQL: asyncpg, psycopg2-binary
  • SQLite: aiosqlite
• Cloud Storage (optional):
  • Azure (Data Lake): chainlit-azure
  • Azure Blob: chainlit-azure-blob
  • Google Cloud: chainlit-gcs
  • AWS S3: chainlit-s3