- Home
- MCP servers
- PostgreSQL Full Access
PostgreSQL Full Access
- typescript
24
GitHub Stars
typescript
Language
4 months ago
First Indexed
3 weeks ago
Catalog Refreshed
Documentation & install
Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.
Installation
Add the following to your MCP client configuration file.
Configuration
View docs{
"mcpServers": {
"syahiidkamil-mcp-postgres-full-access": {
"command": "npx",
"args": [
"-y",
"mcp-postgres-full-access",
"postgresql://username:password@localhost:5432/database"
],
"env": {
"PG_MAX_CONNECTIONS": "20",
"MONITOR_INTERVAL_MS": "5000",
"TRANSACTION_TIMEOUT_MS": "60000",
"PG_STATEMENT_TIMEOUT_MS": "30000",
"ENABLE_TRANSACTION_MONITOR": "true",
"MAX_CONCURRENT_TRANSACTIONS": "5"
}
}
}
}You have a PostgreSQL Full Access MCP Server that lets you read and write to PostgreSQL databases through a controlled, transaction-managed interface. It supports rich schema details, safety features, and explicit commit/rollback steps so you can confidently modify data while keeping guardrails in place.
How to use
Connect your MCP client to the PostgreSQL Full Access MCP Server to start performing data tasks. You can issue read queries, data modifications, and schema changes with built-in safety workflows. For any operation that changes data or structure, the system runs within a transaction and requires your explicit confirmation to commit, giving you multiple opportunities to review changes before they become permanent.
How to install
Prerequisites you need before installing are Node.js and a PostgreSQL database you can access. You will run a local MCP client that communicates with the server via MCP configuration.
Install the MCP client package and set up your environment according to the instructions below. The configuration example assumes you will start the MCP client with a runtime command that invokes the MCP server.
Additional sections
Configuration: You control how the MCP client connects to the server and how long transactions may run. See the example runtime configuration for the client below.
Security: Always operate with a database user that has the minimum necessary permissions. Use a testing database for exploration and review write operations in a dedicated sandbox before applying changes to production data.
Troubleshooting and notes: If a transaction times out or a long-running query is detected, the system can automatically roll back to prevent unintended changes. Review your connection details, transaction timeouts, and user permissions if you encounter issues.
Available tools
execute_query
Execute read-only SQL queries (SELECT). Runs in a READ ONLY transaction and returns execution time and field info.
execute_dml_ddl_dcl_tcl
Execute data modification or schema change statements (INSERT, UPDATE, DELETE, CREATE, ALTER, DROP, etc.). Returns a transaction ID for explicit commit and ends the conversation to review results.
execute_maintenance
Run maintenance commands like VACUUM or ANALYZE or CREATE DATABASE outside of transactions. Provides execution time metrics.
execute_commit
Explicitly commit a transaction by its ID to permanently apply changes to the database.
execute_rollback
Explicitly rollback a transaction by its ID to discard changes.
list_tables
List all tables in the database with column counts and descriptions when available.
describe_table
Provide detailed schema information for a specified table, including keys, indexes, and column metadata.