Use Cases
Common Use Cases
- Natural language to SQL query conversion
- Automated data analysis and reporting
- Database schema documentation generation
- Data quality auditing
Before You Begin
Prerequisites
- PostgreSQL database with connection credentials
- Anthropic API key
- Node.js 18+ installed
Walkthrough
Step-by-Step Guide
1
Install Dependencies
Install the PostgreSQL MCP server and Anthropic SDK.
npm install @modelcontextprotocol/server-postgres @anthropic-ai/sdk2
Connect to Database
Initialize the PostgreSQL MCP server with your connection string.
3
Build the Query Agent
Create an agent that translates natural language to SQL and executes queries.
4
Add Safety Guards
Implement query validation to prevent destructive operations.
Examples
Code Examples
javascript
Natural Language Query Agent
async function queryWithNaturalLanguage(question) {
// Get schema context
const tables = await postgres.tool("list_tables");
const schemas = await Promise.all(
tables.map(t => postgres.tool("describe_table", { table: t }))
);
// Generate SQL
const sql = await claude.generate({
prompt: `Given these tables: ${JSON.stringify(schemas)}
Convert this question to SQL: ${question}
Return ONLY the SQL query.`
});
// Execute and format results
const results = await postgres.tool("query", { sql });
return claude.generate({
prompt: `Format these results as a clear answer: ${JSON.stringify(results)}`
});
}Help
Troubleshooting
Connection refused errors?+
Queries returning unexpected results?+
Quick Info
DifficultyIntermediate
Time Estimate4-5 hours
Tools
PostgreSQLClaude