mcp-postgres
PostgreSQL administration MCP server for read-only inspection, monitoring, and non-destructive maintenance, enabling safe database interaction via assistants like Claude without data modification.
README
mcp-postgres
Serveur MCP (Model Context Protocol) d'administration PostgreSQL — périmètre lecture + maintenance non destructive. Conçu pour brancher un assistant (Claude Code, Claude Desktop, IDE) sur une base afin de l'inspecter, la monitorer et lancer des opérations de maintenance sûres, sans jamais pouvoir écrire de données ni altérer la structure.
Périmètre
| Autorisé | Interdit (aucun tool exposé) |
|---|---|
| Introspection du schéma | DROP, TRUNCATE, DELETE, UPDATE, INSERT |
| Monitoring / activité / verrous | ALTER SYSTEM, gestion de rôles, GRANT/REVOKE |
Statistiques & plans (EXPLAIN, sans ANALYZE) |
VACUUM FULL, pg_terminate_backend |
SELECT ad hoc borné (LIMIT forcé) |
Tout SQL modifiant des données |
VACUUM / ANALYZE / REINDEX CONCURRENTLY (opt-in) |
Défense en profondeur
- Rôle SQL à moindre privilège (deploy/roles/mcp_readonly_role.sql) — rempart final.
- Séparation physique des connexions lecture (READ ONLY) et maintenance (src/mcp_postgres/db/session.py).
- Validateur SQL par vrai parseur (
pglast), pas de regex (src/mcp_postgres/db/guard.py). - Maintenance opt-in (
MCP_MAINTENANCE_ENABLED=falsepar défaut),statement_timeout,LIMITplafonné, secrets jamais loggés.
Installation
python -m venv .venv && . .venv/Scripts/activate # Windows: .venv\Scripts\Activate.ps1
pip install -e ".[dev]"
cp .env.example .env # puis renseigner les identifiants d'un rôle à moindre privilège
Lancement
# stdio (défaut) — pour un client MCP local
mcp-postgres
# ou en HTTP
MCP_TRANSPORT=http mcp-postgres
Intégration Claude Code / Desktop (stdio)
{
"mcpServers": {
"postgres": {
"command": "mcp-postgres",
"env": { "PG_HOST": "localhost", "PG_USER": "mcp_readonly", "PG_PASSWORD": "..." }
}
}
}
Tests
pytest # unitaires (guard, config) sans base
pytest tests/integration # intégration (nécessite Docker via testcontainers)
Structure
Voir docs/capabilities.md (carte des tools/resources/prompts) et docs/security.md (points de vigilance).
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.