SQLite Skill

Local SQLite database operations using MCP server.

When to Use

• •Local database queries and operations
• •Prototyping before moving to production database
• •Data analysis on SQLite files
• •Schema exploration and documentation
• •Test database setup and teardown

Configuration

Specify the database path in the args:

mcp:
  sqlite:
    command: npx
    args: ["-y", "@anthropic/mcp-server-sqlite", "--db-path", "./path/to/database.db"]

For in-memory database:

mcp:
  sqlite:
    command: npx
    args: ["-y", "@anthropic/mcp-server-sqlite", "--db-path", ":memory:"]

Available Tools

Query Operations

• •read_query - Execute SELECT queries (safe, read-only)
• •write_query - Execute INSERT/UPDATE/DELETE queries
• •create_table - Create new tables with schema

Schema Operations

• •list_tables - List all tables in the database
• •describe_table - Get table schema (columns, types, constraints)

Data Operations

• •insert_rows - Bulk insert data
• •export_query - Export query results

Usage Pattern

1. •List tables first: Use list_tables to understand database structure
2. •Describe before querying: Use describe_table for column names and types
3. •Use read_query for exploration: Safe, cannot modify data
4. •Backup before writes: Always backup important data before modifications

Example Workflow

1. list_tables() -> see available tables
2. describe_table(table_name="users") -> see columns
3. read_query(query="SELECT * FROM users LIMIT 10") -> preview data
4. write_query(query="UPDATE users SET status = 'active' WHERE id = 1")

Notes

• •Database file is created if it doesn't exist
• •In-memory databases are lost when server stops
• •Use --db-path to point to existing SQLite files
• •Supports standard SQLite syntax and functions