wpostgresql-mcp

MCP Server for WPostgreSQL Architecting — teaches AI agents how to build high-performance PostgreSQL-backed systems with the wpostgresql library.

Features

• 10 Architect Tools: Blueprints, manual, catalog search, scaffolding, connection validation, pattern generation
• 22 Offline Patterns: Core CRUD, async, batch, transaction, pooling, schema sync, wsqlite backup, and Forensic Audit (ForensicModel soft delete & status=99)
• Scaffolding Generator: Deploys full project structure with config, models, repositories, migrations, main.py
• Live Connection Validation: Tests PostgreSQL connectivity via psycopg 3 and returns health status
• Pattern-based Generation: Generates starter projects from catalog patterns

Relevant Technologies & Key Libraries

• Model Context Protocol (MCP): mcp[cli]>=1.0.0 for AI agent interaction
• Python 3.10+: Type hints and async I/O
• wpostgresql: Type-safe PostgreSQL ORM with Pydantic integration and ForensicModel support
• Pydantic v2: Data model validation and schema generation

Installation

# Clone and install
cd wpostgresql-mcp
pip install -e .

Or use the installer:

chmod +x installer.sh
./installer.sh

Quick Start

# Run the MCP server in stdio mode (for agent integration)
wpostgresql-mcp run

# Get configuration for your AI agent
wpostgresql-mcp config

# Start as background SSE server
wpostgresql-mcp start

Available Tools

Agent Integration

Gemini CLI

gemini mcp add wpostgresql-mcp python3 -m wpostgresql_mcp.server run

Claude Desktop / Cursor

{
  "mcpServers": {
    "wpostgresql-mcp": {
      "command": "python3",
      "args": ["-m", "wpostgresql_mcp.server", "run"]
    }
  }
}

OpenCode

Add to .opencode/config.json:

{
  "mcpServers": {
    "wpostgresql-mcp": {
      "command": "python3",
      "args": ["-m", "wpostgresql_mcp.server", "run"]
    }
  }
}

Commands

Generated by WPostgreSQL MCP by wisrovi