PostgreSQL MCP Server

🏆 Enterprise-Grade Features

📊 Complete Tool Suite

PostgreSQL MCP Server provides 63 specialized tools across 9 comprehensive categories:

• Core Database Operations (9 tools) - Schema management and SQL execution
• JSON Operations (11 tools) - JSONB validation and transformation
• Text Processing (5 tools) - Using pg_trgm and fuzzystrmatch
• Statistical Analysis (8 tools) - Correlation and regression analysis
• Performance Intelligence (6 tools) - AI-powered optimization via hypopg
• Vector/Semantic Search (8 tools) - pgvector integration for AI workflows
• Geospatial Operations (7 tools) - PostGIS support for location data
• Backup & Recovery (4 tools) - Disaster recovery strategies
• Monitoring & Alerting (5 tools) - Real-time health checks

✨ Intelligent Database Assistant

PostgreSQL MCP Server transforms from a tool collection into an intelligent database expert with real-time meta-awareness and guided workflows:

🔮 MCP Resources (10) - Database Meta-Awareness

Real-time database context that AI can access automatically without explicit tool calls:

• database://schema - Complete database structure with tables, columns, indexes
• database://capabilities - Server features, extensions, pg_stat_statements status
• database://performance - Top queries, cache hit rates, slow query detection
• database://health - Connection pool, indexes, vacuum status
• database://extensions - pgvector, PostGIS, hypopg availability with versions
• database://indexes - Usage statistics, size, recommendations
• database://connections - Active connections and pool utilization
• database://replication - Replication lag and health monitoring
• database://vacuum - Bloat detection and transaction ID wraparound
• database://locks - Active locks and blocking query detection

Key Benefit: Resources reduce token usage by providing cached context instead of repeated queries, enabling AI to understand your database state proactively.

🎯 MCP Prompts (10) - Guided Workflows

Step-by-step workflows for complex database operations with PostgreSQL best practices:

• optimize_query - Complete query optimization workflow
• index_tuning - Comprehensive index analysis and recommendations
• database_health_check - Full health assessment with actionable insights
• setup_pgvector - Complete pgvector installation and configuration guide
• setup_postgis - Complete PostGIS setup and spatial query guide
• json_operations - JSONB best practices and optimization strategies
• performance_baseline - Establish performance monitoring baselines
• backup_strategy - Design enterprise backup and recovery strategies
• extension_setup - Step-by-step extension installation guide
• explain_analyze_workflow - Deep dive into query execution plans

Key Benefit: Prompts guide users through multi-step operations with PostgreSQL-specific best practices, reducing errors and improving outcomes.

🔄 Latest Updates (v1.2.0)

Version 1.2.0 introduces intelligent tool filtering to optimize the server for any MCP client:

• 🎛️ Tool Filtering System - New POSTGRES_MCP_TOOL_FILTER environment variable controls which tools are exposed. Use -vector,-geo to disable extension-dependent tools, or -vector,-geo,-stats,-text for Windsurf's 100-tool limit.
• 💰 Token Savings (24-86%) - Reduce token overhead by filtering unused tool groups. Core-only configuration saves ~10,800 tokens per conversation, translating to $3.24 saved at GPT-4 pricing.
• 🎯 9 Tool Groups - Organize 63 tools into core, json, text, stats, performance, vector, geo, backup, and monitoring groups for flexible configuration.
• ✅ Client Compatibility - Stay under Windsurf's 100-tool hard limit and avoid Cursor's ~80 tool performance warnings while maintaining full functionality.
• 🔧 Flexible Syntax - Use -group to disable all tools in a group, -tool to disable specific tools, or +tool to re-enable after a group disable.

Backward Compatibility: v1.2.0 is a drop-in replacement with zero breaking changes. Without filtering, all 63 tools remain enabled.

🚀 Quick Start

Docker (Recommended)

Pull the latest image:

docker pull writenotenow/postgres-mcp-enhanced:latest

Run with your database connection:

docker run -i --rm \
  -e DATABASE_URI="postgresql://user:pass@localhost:5432/db" \
  writenotenow/postgres-mcp-enhanced:latest \
  --access-mode=restricted

Python Installation

pip install postgres-mcp-enhanced

Run the server:

postgres-mcp --access-mode=restricted

From Source

git clone https://github.com/neverinfamous/postgres-mcp.git
cd postgres-mcp
uv sync
uv run pytest -v

🔒 Security Features

Security is a top priority in PostgreSQL MCP Server. We've implemented comprehensive protection including SQL injection prevention through parameter binding, dual security modes for production and development, advanced query validation, 20+ security test cases, CodeQL scanning, and zero known vulnerabilities with a clean security audit.

For detailed security implementation examples and best practices, see our Security Best Practices Gist.

📖 MCP Configuration

Configure for Claude Desktop or Cursor by adding to your MCP settings file:

{
  "mcpServers": {
    "postgres": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "DATABASE_URI",
        "-e", "POSTGRES_MCP_TOOL_FILTER",
        "writenotenow/postgres-mcp-enhanced:latest",
        "--access-mode=restricted"
      ],
      "env": {
        "DATABASE_URI": "postgresql://user:pass@localhost:5432/db",
        "POSTGRES_MCP_TOOL_FILTER": "-vector,-geo"
      }
    }
  }
}

The POSTGRES_MCP_TOOL_FILTER is optional. Use -vector,-geo to disable extension-dependent tools, or -vector,-geo,-stats,-text for clients with tool limits like Windsurf.

💡 Use Cases & Expert Guides

PostgreSQL MCP Server excels in diverse scenarios including AI-native data operations with pgvector embeddings and semantic search, location-based services with PostGIS geospatial queries, performance engineering with automated index recommendations and query optimization, plus comprehensive data analytics with statistical analysis and time series processing.

Master PostgreSQL MCP operations with comprehensive expert guides:

💡 Pro Tip: These expert guides include production-ready code examples and best practices. Start with the Complete Feature Showcase to understand the full scope, then dive into specialized guides for your specific use case.

📚 Resources

Find comprehensive documentation, source code, container images, and installation packages through our official channels: Documentation Wiki, GitHub Repository, Docker Hub, and PyPI Package.

🎯 What's Next

PostgreSQL MCP Server represents a major milestone in bringing enterprise-grade database operations to AI-native workflows. With 63 specialized tools, 10 intelligent resources for database meta-awareness, 10 guided prompts for complex operations, zero known vulnerabilities, and production-ready reliability, it's ready to transform how you work with PostgreSQL databases.

The intelligent assistant capabilities enable AI to understand your database context proactively, provide PostgreSQL-specific recommendations, and guide you through complex operations with best practices built in.

💡 Pro Tip: Start with restricted mode for production environments, explore the MCP Resources for instant database insights, use MCP Prompts for guided workflows, and leverage the 63 specialized tools to streamline your database operations.

🔗 Related Resources

Explore our MCP Server Hub for additional tools including SQLite MCP Server and Memory Journal MCP.

🏆 Why PostgreSQL MCP Matters

PostgreSQL MCP delivers production-ready reliability with zero known vulnerabilities, intelligent database assistant with 10 real-time resources and 10 guided prompts, AI-native operations through pgvector support, performance intelligence with hypothetical index testing, geospatial capabilities via PostGIS, 63 specialized tools across 9 categories, intelligent tool filtering for client compatibility and token savings, and a security-first design with dual access modes. This represents the evolution from a powerful tool collection to a truly intelligent database expert.