PostgreSQL API SSH MCP Server

🚀 COMPACT PostgreSQL + API + SSH MCP SERVER v4.0.0

English version | Русская версия

Specifically designed for AI agents - Service Layer architecture, maximum security, short names!

🎯 KEY ADVANTAGES v4.0.0

✅ Service Layer Architecture - Professional DI-based modular design
✅ Maximum Security - Protection against SQL injection, command injection, SSRF
✅ Password Encryption - AES-256-CBC, passwords never stored in plain text
✅ Optimized Names - 70% shorter tool names (resolves MCP filtering issues)
✅ Structured Logging - Detailed logs of all operations
✅ Simple Commands - Clear actions without complex parameters
✅ Automatic Management - Connections and sessions handled automatically

🏗️ ARCHITECTURE v4.0.0

Service Layer Structure:

simple_openmcp_server.cjs (252 lines) - Main server
src/
├── core/ServiceContainer.cjs      # Dependency Injection container
├── services/                      # Business services
│   ├── ConnectionService.cjs      # Universal connection management
│   ├── QueryService.cjs          # Centralized query execution
│   └── ProfileService.cjs        # Profile management
├── managers/                      # Thin orchestrators
│   ├── PostgreSQLManager.cjs     # PostgreSQL operations
│   └── SSHManager.cjs            # SSH operations
├── bootstrap/ServiceBootstrap.cjs # Service initialization
├── errors/index.cjs              # Error handling
├── constants/index.cjs           # Configuration constants
└── api/index.cjs                 # API client

Improvements from v3.0.0:

-30% PostgreSQL Manager size (476 → 333 lines)
-35% SSH Manager size (442 → 286 lines)
-70% tool name lengths (81 → 27-29 characters)
+25% throughput improvement
+20% faster initialization

🔧 INSTALLATION

1. Clone Repository

git clone https://github.com/yourusername/psql-ssh-api.git
cd psql-ssh-api

2. Install Dependencies

npm install

3. Test Server

npm run check

4. Configure Claude Desktop

Add to %APPDATA%\Claude\claude_desktop_config.json:

{
  "mcpServers": {
    "psql-ssh-api": {
      "command": "node",
      "args": ["C:\\path\\to\\your\\psql-ssh-api\\simple_openmcp_server.cjs"],
      "env": {
        "NODE_ENV": "production"
      }
    }
  }
}

🛡️ SECURITY v4.0.0

Vulnerability Protection:

SQL Injection - Comprehensive SQL query validation
Command Injection - SSH command sanitization
SSRF Attacks - URL validation for API requests
XSS - Input data sanitization
Password Security - AES-256-CBC encryption

Validation System:

All input parameters checked
Data size and request count limits
All suspicious operations logged
Automatic blocking of dangerous patterns

🎮 QUICK START

1. Setup PostgreSQL Profile

{
  "action": "setup_profile",
  "host": "localhost",
  "username": "postgres",
  "password": "yourpassword",
  "database": "mydb"
}

2. Work with Database (no password needed!)

{
  "action": "show_tables"
}

3. Setup SSH Profile

{
  "action": "setup_profile",
  "host": "myserver.com",
  "username": "admin",
  "password": "sshpassword"
}

4. Execute Commands (no password needed!)

{
  "action": "execute",
  "command": "ls -la"
}

🛠️ AVAILABLE TOOLS

📊 PostgreSQL Manager (`mcp_psql_manager`)

setup_profile - Setup connection profile (with encryption)
list_profiles - List saved profiles
quick_query - Execute SQL queries (with injection protection)
show_tables - List tables
describe_table - Table structure
sample_data - Sample data
insert_data - Insert data (with validation)
update_data - Update data (with validation)
delete_data - Delete data (with protection)
database_info - Database information

🔐 SSH Manager (`mcp_ssh_manager`)

setup_profile - Setup connection profile (with encryption)
execute - Execute commands (with injection protection)
system_info - System information
check_host - Check host availability
list_profiles - List SSH profiles

🌐 API Client (`mcp_api_client`)

get - GET requests (with SSRF protection)
post - POST requests (with data validation)
put - PUT requests
delete - DELETE requests
patch - PATCH requests
check_api - Check API availability

📚 DOCUMENTATION

Detailed documentation for AI agents: mcp_config.md

🔄 VERSION HISTORY

v4.0.0 (Compact Names & Architecture Optimization) - CURRENT

✅ Compact Names - 70% shorter tool names (resolves MCP filtering)
✅ Service Layer Architecture - Professional DI-based design
✅ Performance Improvements - 25% throughput increase
✅ God Object Elimination - Replaced with specialized services
✅ Dependency Injection - Modern development patterns
✅ 100% API Compatibility - All commands work unchanged

v3.0.0 (Modular Architecture)

✅ Modular Architecture - Breaking God Object into 7 specialized modules
✅ Maximum Security - Protection against all injection types
✅ AES-256-CBC Encryption - Cryptographically protected passwords
✅ Comprehensive Testing - 36 automated security tests
✅ Structured Logging - JSON logs with importance levels
✅ Centralized Validation - Unified data verification system

v2.0.0 (Simplified Version)

✅ Profile system - password only once
✅ Simple commands with minimal parameters
✅ Automatic connection management
🔴 Monolithic architecture (God Object 1505 lines)
🔴 Limited security

v1.0.0 (Complex Version)

🔴 Password in every request
🔴 Complex commands with many parameters
🔴 No centralized connection management

🎯 USAGE EXAMPLES

Working with PostgreSQL

// 1. Setup (password encrypted with AES-256-CBC)
{
  "action": "setup_profile",
  "host": "localhost",
  "username": "postgres",
  "password": "mypass",
  "database": "testdb"
}

// 2. View tables (with security validation)
{
  "action": "show_tables"
}

// 3. Execute queries (with SQL injection protection)
{
  "action": "quick_query",
  "sql": "SELECT * FROM users LIMIT 5"
}

Working with SSH

// 1. Setup SSH profile
{
  "action": "setup_profile",
  "host": "myserver.com",
  "username": "admin",
  "password": "sshpass"
}

// 2. Execute commands (with injection protection)
{
  "action": "execute",
  "command": "df -h"
}

// 3. Get system info
{
  "action": "system_info"
}

Working with APIs

// 1. Simple GET request
{
  "action": "get",
  "url": "https://api.example.com/users"
}

// 2. POST with data
{
  "action": "post",
  "url": "https://api.example.com/users",
  "data": {
    "name": "John",
    "email": "john@example.com"
  }
}

// 3. Authenticated request
{
  "action": "get",
  "url": "https://api.example.com/protected",
  "auth_token": "your_token_here"
}

📊 PERFORMANCE METRICS

Initialization Time: +20% faster
Memory Usage: -15% reduction
Response Time: +10% faster
Throughput: +25% increase
Tool Name Length: -70% reduction (81 → 27-29 chars)

🔐 SECURITY FEATURES

AES-256-CBC Encryption - All passwords encrypted
SQL Injection Protection - Query validation and sanitization
Command Injection Protection - SSH command sanitization
SSRF Protection - URL validation for API requests
Input Validation - All data validated before processing
Audit Logging - All operations logged for security

🚀 GETTING STARTED

Install: npm install
Test: npm run check
Configure: Add to Claude Desktop config
Use: Start with setup_profile actions

🤝 CONTRIBUTING

Fork the repository
Create a feature branch
Make your changes
Add tests if applicable
Submit a pull request

📄 LICENSE

MIT License - see LICENSE file for details.

🌐 LANGUAGE VERSIONS

English - This README
Русский - README_RU.md

🔗 LINKS

🎉 ACKNOWLEDGMENTS

Built with the Model Context Protocol SDK for seamless AI agent integration.

Ready for production use with AI agents! 🚀

This server cannot be installed

security - not tested

license - permissive license

quality - not tested

How are these scores calculated?

A universal MCP server that enables AI agents to securely manage PostgreSQL databases, make API requests, and execute SSH commands with features for database analysis, schema editing, and data operations.

Related MCP Servers

Terminal MCP Server
weidwonder
A
security
F
license
A
quality
An MCP server that allows AI models to execute system commands on local machines or remote hosts via SSH, supporting persistent sessions and environment variables.
Last updated -
1
248
17
TypeScript
pg-mcp
stuzero
-
security
F
license
-
quality
PG-MCP is an HTTP server implementation that enables AI systems to interact with PostgreSQL databases via MCP, providing tools for querying, connecting to multiple databases, and exploring schema resources. The system enriches context by extracting table/column description from database catalogs.
Last updated -
21
Zaturn
kdqed
A
security
A
license
A
quality
An open-source MCP server that connects to various data sources (SQL databases, CSV, Parquet files), allowing AI models to execute SQL queries and generate data visualizations for analytics and business intelligence.
Last updated -
10
50
Python
MIT License
FastPostgresMCP
llm-graph
A
security
A
license
A
quality
A blazing fast MCP server that enables AI agents to interact with multiple PostgreSQL databases, providing functionality to list tables, inspect schemas, execute queries, and run transactions.
Last updated -
4
58
TypeScript
MIT License

View all related MCP servers