Update env.local.example

README.md

@@ -2,6 +2,12 @@
 
 oss-data-analyst is an intelligent AI agent that converts natural language questions into SQL queries and provides data analysis. Built with the Vercel AI SDK, it features multi-phase reasoning (planning, building, execution, reporting) and streams results in real-time.
 
+> **Note**: This is a fork of the original [vercel-labs/oss-data-analyst](https://github.com/vercel-labs/oss-data-analyst) template with modifications to make it more suitable for demonstration purposes. Key changes include:
+>
+> - **Emphasis on local databases**: Optimized for SQLite-based demos without requiring cloud database setup
+> - **Prompt engineering improvements**: Enhanced system prompts for better query understanding and result interpretation
+> - **Streamlined setup**: Simplified configuration and dependencies for quick local deployment
+
 > **Note**: This is a reference architecture. The semantic catalog and schemas included are simplified examples for demonstration purposes. Production implementations should use your own data models and schemas.
 
 ## Features
@@ -25,29 +31,36 @@ oss-data-analyst is an intelligent AI agent that converts natural language quest
 ### Installation
 
 1. **Clone the repository**
+
    ```bash
    git clone https://github.com/vercel/oss-data-analyst.git
    cd oss-data-analyst
    ```
 
 2. **Install dependencies**
+
    ```bash
    pnpm install
    ```
 
 3. **Set up environment variables**
+
    ```bash
    cp env.local.example .env.local
    ```
+
    Edit `.env.local` and add your Vercel AI Gateway key
 
 4. **Initialize the database**
+
    ```bash
    pnpm initDatabase
    ```
+
    This creates a SQLite database with sample data (Companies, People, Accounts)
 
 5. **Run the development server**
+
    ```bash
    pnpm dev
    ```
@@ -67,43 +80,51 @@ pnpm start
 This repository includes a sample database schema with three main entities to demonstrate oss-data-analyst's capabilities:
 
 ### **Companies**
+
 Represents organizations in your database. Each company has:
+
 - Basic information (name, industry, employee count)
 - Business metrics (founded date, status)
 - Example: Technology companies, Healthcare organizations, etc.
 
 ### **Accounts**
+
 Represents customer accounts or subscriptions tied to companies. Each account includes:
+
 - Account identification (account number, status)
 - Financial metrics (monthly recurring value, contract details)
 - Relationship to parent company
 - Example: Active subscriptions with monthly values ranging from $10k-$50k
 
 ### **People**
+
 Represents individual employees or contacts within companies. Each person has:
+
 - Personal information (name, email)
 - Employment details (department, title, salary)
 - Relationship to their company
 - Example: Engineers, Sales representatives, Managers across different departments
 
-
 ## How It Works
 
 oss-data-analyst uses a multi-phase agentic workflow:
 
 1. **Planning Phase**
+
    - Analyzes natural language query
    - Searches semantic catalog for relevant entities
    - Identifies required data and relationships
    - Generates execution plan
 
 2. **Building Phase**
+
    - Constructs SQL query from plan
    - Validates syntax and security policies
    - Optimizes query structure
    - Finds join paths between tables
 
 3. **Execution Phase**
+
    - Estimates query cost
    - Executes SQL against database
    - Handles errors with automatic repair
@@ -120,6 +141,7 @@ oss-data-analyst uses a multi-phase agentic workflow:
 ### Customizing Prompts
 
 Modify system prompts in `src/lib/prompts/`:
+
 - `planning.ts` - Planning phase behavior
 - `building.ts` - SQL generation logic
 - `execution.ts` - Query execution handling
@@ -136,25 +158,28 @@ Try asking oss-data-analyst (using the sample database):
 - "What is the total revenue for Active accounts?"
 - "How many people work in Engineering?"
 
-## Using with Production Databases
+## Using with Your Own Data
 
-The default setup uses SQLite for demonstration. To use with Snowflake or other databases:
+The application uses SQLite as its database. To use your own data:
 
-1. Update `src/lib/oss-data-analyst-agent-advanced.ts` to import from `./tools/execute` instead of `./tools/execute-sqlite`
-2. Configure your database credentials in `.env.local`
-3. Update the semantic catalog in `src/lib/semantic/` with your schema definitions
+1. Modify the database seed script at `scripts/seed-database.ts` with your data
+2. Run `pnpm initDatabase` to reinitialize the database
+3. Update the semantic catalog in `src/semantic/` with your schema definitions matching your data structure
 
 ## Troubleshooting
 
 **Database Not Found**
+
 - Run `pnpm initDatabase` to create and seed the database
 - Check that `data/oss-data-analyst.db` exists
 
 **AI Gateway API Errors**
+
 - Verify your API key is valid in `.env.local`
 - Check API rate limits and credits
 
 **Build Errors**
+
 - Run `pnpm install` to update dependencies
 - Check TypeScript errors with `pnpm run type-check`
 - Clear `.next` folder and rebuild

env.local.example

@@ -6,20 +6,14 @@ NEXT_PUBLIC_API_URL=http://localhost:3000
 # Development
 NODE_ENV=development
 
-# AI Provider (OpenAI)
-OPENAI_API_KEY=your_openai_api_key_here
+# AI Provider
+# Create an API key: https://vercel.com/d?to=%2F%5Bteam%5D%2F%7E%2Fai%2Fapi-keys
+AI_GATEWAY_API_KEY=your_ai_gateway_api_key_here
 
-# Snowflake Database Configuration
-SNOWFLAKE_ACCOUNT=your_account.region
-SNOWFLAKE_USERNAME=your_username
-SNOWFLAKE_PASSWORD=your_password
-# OR use private key authentication (optional)
-# SNOWFLAKE_PRIVATE_KEY=your_private_key_here
-
-SNOWFLAKE_ROLE=your_role
-SNOWFLAKE_WAREHOUSE=your_warehouse
-SNOWFLAKE_DATABASE=your_database
-SNOWFLAKE_SCHEMA=your_schema
+# SQLite Database Configuration
+# The application uses a local SQLite database located at data/oss-data-analyst.db
+# Run 'pnpm run initDatabase' to initialize and seed the database
 
 # Security - Allowed schemas for queries (comma-separated)
-ALLOWED_SCHEMAS=analytics,crm
+# SQLite uses "main" as the default schema
+ALLOWED_SCHEMAS=main

package.json

@@ -94,7 +94,7 @@
     "lucide-react": "^0.468.0",
     "motion": "^12.23.24",
     "nanoid": "^5.1.6",
-    "next": "^15.0.3",
+    "next": "15.0.5",
     "next-themes": "^0.4.6",
     "node-sql-parser": "^5.3.10",
     "pg": "^8.16.3",
@@ -106,8 +106,6 @@
     "react-resizable-panels": "^3.0.6",
     "react-syntax-highlighter": "^15.6.6",
     "react-typed": "^2.0.12",
-    "recharts": "2.15.4",
-    "snowflake-sdk": "^2.1.3",
     "sonner": "^2.0.7",
     "streamdown": "^1.4.0",
     "tailwind-merge": "^3.3.1",
@@ -116,8 +114,6 @@
     "tw-animate-css": "^1.4.0",
     "use-stick-to-bottom": "^1.1.1",
     "vaul": "^1.1.2",
-    "vega": "^6.2.0",
-    "vega-lite": "^6.4.1",
     "zod": "^4.1.12"
   },
   "packageManager": "pnpm@8.15.0",