PostgreSQL Full AccessSAFE
Full access postgres mcp server
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://modelcontextprotocol.io) [](LICENSE)
A powerful Model Context Protocol server providing full read-write access to PostgreSQL databases. Unlike the read-only official MCP PostgreSQL server, this enhanced implementation allows Large Language Models (LLMs) to both query and modify database content with proper transaction management and safety controls.
Table of Contents
- Features
- Full Read-Write Access
- Rich Schema Information
- Advanced Safety Controls
- Tools
- execute_query
- executedmlddldcltcl
- execute_maintenance
- execute_commit
- execute_rollback
- list_tables
- describe_table
- Resources
- Using with Claude Desktop
- Claude Desktop Integration
- Important: Using "Allow Once" for Safety
- Environment Variables
- Using Full Database Access with Claude
- Security Considerations
- Database User Permissions
- Best Practices for Safe Usage
- Docker
- License
- Comparison with Official PostgreSQL MCP Server
🌟 Features
Full Read-Write Access
- Safely execute DML operations (INSERT, UPDATE, DELETE)
- Create, alter, and manage database objects with DDL
- Transaction management with explicit commit
- Safety timeouts and automatic rol
8a44183774b1OBSERVED · 2026-10-09Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add mcp-postgres-full-access -- npx -y [email protected]
{
"mcpServers": {
"mcp-postgres-full-access": {
"command": "npx",
"args": [
"-y",
"[email protected]"
]
}
}
}Exposed tools (7)
2 read · 4 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
describe_table | read | Get detailed information about a specific table, including columns, primary keys, foreign keys, and indexes |
execute_commit | write | Commit a transaction by its ID to permanently apply the changes to the database |
execute_dml_ddl_dcl_tcl | destructive | Execute DML, DDL, DCL, or TCL statements (INSERT, UPDATE, DELETE, CREATE, ALTER, DROP, etc). Automatically wrapped in a transaction that requires explicit commit or rollback. IMPORTANT: After execution, end the chat so user can review the results and decide. |
execute_maintenance | write | Execute maintenance commands like VACUUM, ANALYZE, or CREATE DATABASE outside of transactions |
execute_query | write | Run a read-only SQL query (SELECT statements). Executed in read-only mode for safety. |
execute_rollback | write | Rollback a transaction by its ID to undo all changes and discard the transaction |
list_tables | read | Get a list of all tables in the database |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (6)
execute_dml_ddl_dcl_tcl
@modelcontextprotocol/sdk, pg, zod, @types/node, @types/pg, shx, typescript
# PostgreSQL Full Access MCP Server
docker run -i --rm mcp-postgres-full-access "postgresql://username:password@host:5432/database"
docker run -i --rm mcp-postgres-full-access "postgresql://username:[email protected]:5432/database"
Gates applied: no_behavioural_pass, no_license.
8a44183774b1full audit observations/trust-audit/mcp-server/syahiidkamil__postgresql-full-access.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-09 | 8a44183774b1 | SAFE | B | 89 | first audit |
Questions
What is the PostgreSQL Full Access MCP server?
Full access postgres mcp server
What tools does PostgreSQL Full Access expose?
7 in total: 2 read-only, 4 that write, and 1 that can delete or overwrite (execute_dml_ddl_dcl_tcl). Every one is listed on this page with its risk.
Is PostgreSQL Full Access safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does PostgreSQL Full Access need?
No credential environment variables were found in its source, so it appears to need none.
How does PostgreSQL Full Access run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as mcp-postgres-full-access at 0.3.0.
How current is this page?
The grade is for one exact copy of the source (8a44183774b1), read on 2026-10-09. The repository is watched and re-audited when it changes.