Show HN: PostgreSQL MCP Server with 135 tools for various purpose

A developer released mcp-postgres, a high-performance MCP server written in Rust that provides 135 PostgreSQL tools for AI tools like Claude Desktop, featuring lock-free connection pooling and sub-10ms latency. The server supports TCP, HTTP/2, and stdio transports, and implements the Model Context Protocol revision 2025-11-25.

mcp-postgres is a high-performance MCP server that brings PostgreSQL into Claude Desktop and any MCP-compatible AI tool. 135 PostgreSQL tools, lock-free connection pooling, sub-10ms latency. MCP suite.One of four high-performance MCP servers written in Rust — mcp-postgres · mcp-filesystem · mcp-memory · mcp-web-search . All implement MCP protocol revision. 2025-11-25 From crates.io cargo install mcp-postgres Or from Homebrew macOS brew tap corporatepiyush/mcp-postgres brew install mcp-postgres Stdio mode for Claude Desktop mcp-postgres --database-url "postgres://user:pass@localhost:5432/mydb" --stdio TCP server port 3000 mcp-postgres --database-url "postgres://user:pass@localhost:5432/mydb" HTTP/2 server port 3001 mcp-postgres --database-url "postgres://user:pass@localhost:5432/mydb" --http-port 3001 Add to claude desktop config.json : { "mcpServers": { "postgres": { "command": "mcp-postgres", "args": "--database-url", "postgres://user:pass@localhost:5432/mydb", "--stdio" } } } | Feature | mcp-postgres | DIY / psql | |---|---|---| 135 purpose-built tools | Schema inspection, DDL, monitoring, replication, batch ops, security audit, text search, extensions, maintenance, and more | You build every query from scratch | Lock-free connection pool | Zero-mutex crossbeam::ArrayQueue — pure CAS loops, no kernel overhead | Deadpool or manual Mutex<VecDeque | Dual-protocol | TCP 3000 + HTTP/2 3001 + stdio — one binary, three transports | Multiple servers to wire up | Sub-10ms latency | Allocated for AI interactivity — hot path is allocation-free | Unpredictable | SQL injection prevention | Every identifier validated, quote ident sanitization, structured predicates | Manual parameterization | PG version-aware | Queries verified against PG 16–18 docs, graceful fallbacks for version differences | Version-specific failures | Usage: mcp-postgres OPTIONS Options: -d, --database-url <URL PostgreSQL connection string -H, --host <HOST TCP server host 127.0.0.1 -p, --port <PORT TCP server port 3000 --http-port <PORT HTTP/2 server port 3001 --min-connections <N Min pool connections 5 --max-connections <N Max pool connections 20 --log-level <LEVEL Log level info --enable-metrics Prometheus /metrics endpoint --metrics-port <PORT Metrics port 9090 --stdio Stdio mode Claude Desktop --access-mode <MODE unrestricted, restricted unrestricted -h, --help Print help -V, --version Print version Implements the Model Context Protocol https://modelcontextprotocol.io revision 2025-11-25 over JSON-RPC 2.0 https://spec.modelcontextprotocol.io , via TCP, HTTP/2, or stdio. | Area | Support | |---|---| | Transports | stdio, TCP 3000 , HTTP/2 3001 | | Protocol version | 2025-11-25 , negotiates down to 2025-06-18 / 2025-03-26 / 2024-11-05 | initialize | ✅ version negotiation + instructions | tools/list , tools/call | ✅ 135 tools | CallToolResult | ✅ content + structuredContent + isError | | Capabilities advertised | tools only — nothing is advertised that isn't implemented | resources · prompts · logging · completion | ❌ roadmap — see | Request: { "jsonrpc": "2.0", "method": "tools/call", "params": { "name": "list tables", "arguments": {} }, "id": 1 } Result — a spec-compliant CallToolResult . The payload is available as a machine-readable structuredContent object and as serialized text ; tool failures come back with isError: true not as JSON-RPC protocol errors so the model can self-correct: { "content": { "type": "text", "text": "{\"tables\": \"users\",\"orders\" }" } , "structuredContent": { "tables": "users", "orders" }, "isError": false } Upgrading from 4.x? The result shape changed — see MIGRATION.md . ⚡ Query Execution 8 — execute query, execute insert, execute update, execute delete, explain query, async execute insert, async execute update, async execute delete Fire off raw SQL, run parameterized inserts/updates/deletes with automatic type coercion, and peek under the hood with EXPLAIN ANALYZE plans. Async variants let you fire-and-forget long-running operations without blocking your AI workflow. 🪄 Key moves: execute query for ad-hoc SQL, explain query to spot missing indexes or seq-scans, async execute for bulk writes that outlive the request. 🔍 Schema Inspection 8 — list tables, describe table, list schemas, list indexes, list triggers, show constraints, list partitions, get object details Peel back the layers of your database: list every table across schemas, drill into column types and nullability, inspect index definitions, trigger functions, check constraints, and navigate partitioned table hierarchies. 🪄 Key moves: describe table is your go-to for column metadata, get object details shows DDL + stats in one shot, list partitions maps your partitioning tree. 🏗️ DDL Operations 15 — create/drop table, view, schema, sequence, index, partition, alter view, backup table Full DDL surface for schema evolution. Spin up tables with typed columns, create views to simplify complex queries, generate sequences for auto-increment IDs, build indexes for performance, and partition large tables for manageability. 🪄 Key moves: backup table snapshots a table before risky DDL, alter view redefines without dropping, create partition attaches new ranges to existing partition trees. 📦 Batch Operations 4 — async batch insert, async batch update, async batch delete, async batch insert copy Move mountains of data in a single call. Insert thousands of rows with structured arrays, run bulk updates and deletes with filtered predicates, or use async batch insert copy leveraging PostgreSQL's COPY protocol for wire-speed ingestion. 🪄 Key moves: async batch insert copy is 2-3x faster than row-by-row INSERT for 10K+ rows, async batch update handles conditional multi-row updates atomically. 📊 Database Monitoring 10 — table/index stats, database/table size, cache hit ratio, vacuum, analyze, pg stat statements, reset statistics See how your database is really doing. Track table and index usage stats, measure disk consumption per database or table, calculate cache hit ratios, kick off VACUUM and ANALYZE, and query pg stat statements to find your top CPU-hungry queries. 🪄 Key moves: get cache hit ratio tells you if your shared buffers are sized right, get pg stat statements surfaces slow queries by total time, analyze table refreshes planner stats on-demand. 🔌 Connection Management 4 — list connections, show current user, show running queries, show connection summary See who's connected, what they're running, and how connections are distributed. Diagnose connection bloat, find runaway queries, and identify which application is hogging the pool. 🪄 Key moves: show running queries catches long-running queries in-flight, show connection summary groups by state/application for a bird's-eye view. 🔐 Security & Users 5 — list users, user/role privileges, database privileges, session info Audit your security posture: enumerate database roles and their login capabilities, inspect table-level and schema-level GRANTs, check role membership chains, and review database-level ACLs. 🪄 Key moves: list user privileges surfaces exactly what each user can SELECT/INSERT/UPDATE/DELETE, list role memberships reveals privilege escalation paths through role inheritance. ⚙️ Configuration 5 — all settings, get setting, memory/performance/log settings Navigate the sprawling world of postgresql.conf without grepping. View every GUC parameter, look up specific settings by name, and filter by category to zero in on memory tuning, performance knobs, or logging config. 🪄 Key moves: show memory settings pulls shared buffers, work mem, maintenance work mem in one shot, get setting for a quick SHOW of any parameter. 🔄 Transaction Monitoring 7 — active transactions, locks, waiting locks, isolation, deadlocks, autocommit, transaction timeout Dive into the transaction machinery: detect long-running idle-in-transaction sessions, map lock contention chains with pg blocking pids , identify deadlocks, check isolation levels, and monitor transaction age to prevent bloat. 🪄 Key moves: show waiting locks shows who's blocked by whom, show deadlocks queries pg stat activity for blocked processes, show transaction timeout checks idle in transaction session timeout . 📋 Replication 5 — replication status, replication slots, standby servers, wal info, base backup progress Keep your replicas healthy. Monitor streaming replication lag in bytes and time, inspect replication slots for WAL accumulation, list standby servers with their flush/replay positions, and track pg stat progress basebackup for ongoing backups. 🪄 Key moves: show replication status shows sender/receiver pairs with lag, list replication slots helps you spot slots that are consuming too much WAL. 🏥 Database Health 4 — analyze db health, unused/duplicate indexes, vacuum progress Get a comprehensive wellness check: scan for unused indexes that waste write IO, detect duplicate indexes that bloat storage, monitor VACUUM progress across all databases, and get a health score with actionable recommendations. 🪄 Key moves: list unused indexes finds indexes with zero scans since stats reset, show vacuum progress tracks autovacuum workers in real-time. 🗑️ Maintenance 1 — truncate table Safely and quickly remove all rows from a table while preserving the table structure. Performs privilege validation before executing, and supports RESTART IDENTITY for serial column reset. 🪄 Key moves: Blows away table data faster than DELETE — ideal for staging tables and temp data cleanup. 🧠 Index Advisor 1 — suggest indexes Analyzes your query workload from pg stat statements and suggests index candidates based on WHERE clauses, JOIN conditions, and ORDER BY patterns. Each suggestion includes the estimated impact and DDL to create it. 🪄 Key moves: Run after capturing a representative workload — the suggestions get smarter the more queries pg stat statements has sampled. 🔬 Performance Audit 2 — audit performance, analyze query performance Deep-dive query forensics with buffer-level analysis. Inspect shared hits vs reads, execution time breakdowns by plan node, row estimate accuracy, and temp file spill detection. audit performance runs a full sweep across your top queries. 🪄 Key moves: analyze query performance with EXPLAIN ANALYZE, BUFFERS catches seq-scans on large tables, misestimated row counts, and sort spills to disk. 🛡️ Security Audit 3 — audit security, audit user permissions, audit role hierarchy Run a comprehensive security posture review: find users with superuser privileges, detect excessive schema-level GRANTs, map role inheritance chains that could lead to privilege escalation, and audit columns containing sensitive-sounding names. 🪄 Key moves: audit security produces an executive summary with severity ratings, audit role hierarchy visualizes role grant chains that might bypass intended restrictions. 🔎 Text Search 6 — search vector, levenshtein search, trigram search, soundex search, metaphone search, full text config Unlock fuzzy and phonetic search across your text data. Search by vector similarity via pgvector, find approximate matches with Levenshtein distance, use trigram similarity for flexible substring matching, or try Soundex/Metaphone for phonetic lookups when you don't know the spelling. 🪄 Key moves: trigram search handles typos and partial matches gracefully, soundex search finds names that sound alike, full text config lets you inspect tsvector configuration. 🧩 Extension Management 5 — list extensions, install extension, remove extension, update extension, extension details Manage your PostgreSQL extension ecosystem. List installed extensions with versions, install new extensions from available packages, upgrade to latest versions, remove unused ones, and inspect extension dependencies to understand what cascade effects to expect. 🪄 Key moves: extension details shows extension-to-extension dependency chains, install extension supports CASCADE for dependency auto-install. 📐 Schema Alter 7 — add/drop/alter column, add/drop constraint, set default, drop default Evolve your schema surgically without writing raw DDL. Add nullable or NOT NULL columns with default values, change column types when castable , drop obsolete columns, manage CHECK and UNIQUE constraints, and set or remove column defaults. 🪄 Key moves: add column with NOT NULL + default backfills existing rows, drop default stops future inserts from getting the default without touching existing data. 🕒 Session Management 3 — set session setting, show session settings, reset session setting Twist session-level config knobs without editing postgresql.conf . Uses SET LOCAL scoped to the transaction for safety — settings auto-revert on commit/rollback so you never accidentally leave a modified config behind. 🪄 Key moves: set session setting for per-query statement timeout or work mem tweaks, reset session setting reverts a single setting to its cluster default. 👤 User Management 3 — create user, alter user, drop user Manage database role lifecycle with safety guards. Create users with password and login privileges, alter existing roles rename, change password, toggle superuser/createrole flags , and drop users with ownership reassignment and dependency checks. 🪄 Key moves: drop user refuses if the user owns objects prevents accidental CASCADE carnage , alter user can set per-role GUC defaults like statement timeout . 💾 Data Tools 10 — export table, import table, show table data, search data, compare data, data profile, find duplicates, find orphans, show pg stat user indexes, show table bloat Explore, profile, and clean your data. Export tables to structured results, search across columns with pattern matching, compare two datasets row-by-row, profile column distributions and null ratios, find duplicate rows and orphaned foreign keys, and estimate table bloat to plan VACUUMs. 🪄 Key moves: data profile gives you min/max/null counts/approx distinct per column in one shot, find orphans detects FK violations where parent rows went missing, show table bloat estimates wasted space. ⚗️ Migration Helpers 7 — generate migration, apply migration, rollback migration, list migrations, show migration status, migration history, validate migration End-to-end schema migration workflow. Generate timestamped migration files from DDL templates, apply pending migrations in order, rollback the last applied migration, track which migrations have been run, and validate the current schema state against the migration log. 🪄 Key moves: validate migration detects schema drift manual changes that skipped the migration system , rollback migration reverts safely using generated down-scripts. 📈 Vector Database 1 — search similarity pgvector-powered similarity search for embeddings and vector data. Search by L2 distance, inner product, or cosine similarity with optional metadata filters. Supports indexing IVFFlat, HNSW for fast approximate nearest-neighbor lookups. 🪄 Key moves: Perfect for RAG workflows — feed in your embedding vectors and retrieve the top-K most semantically similar results with a single call. 📆 Time-Series 1 — analyze timescale TimescaleDB integration for time-series workloads. Inspect hypertable chunks, analyze compression ratios, review retention policies, and get recommendations for chunk interval tuning based on your ingestion patterns. 🪄 Key moves: Run analyze timescale to check if your compression policy is keeping up with ingestion volume, or if chunk intervals need resizing. 📂 Data I/O 6 — import csv, export csv, import json, export json, show table data paginated, bulk load from file Move data in and out of PostgreSQL in every common format. Import CSV/JSON with automatic type detection and error handling, export query results to CSV/JSON, paginate through large result sets with keyset pagination, and bulk-load from server-side files. 🪄 Key moves: show table data paginated handles million-row tables without blowing up your context, bulk load from file uses server-side COPY for zero-network-overhead ingestion. No mutexes. No semaphores. Just CAS loops. ┌─────────────┐ ┌──────────────────┐ ┌─────────────┐ │ Task 1 │ │ crossbeam:: │ │ Task 2 │ │ acquire │────▶│ ArrayQueue │◀────│ release │ └─────────────┘ │ CachePadded │ └─────────────┘ │ Head/Tail │ │ CAS only │ └──────────────────┘ | Metric | Deadpool v3.x | LockFreePool v4.0.0 | |---|---|---| | Lock acquisitions per acquire+release | 3+ Mutex + Semaphore | 0 | | Allocation on hot path | Yes VecDeque growth | Zero pre-allocated | | False sharing | Likely adjacent fields | Cache-padded atomics | | Inlining | Cross-crate, opaque | Monomorphic, LTO-friendly | | Dependencies | 2 deadpool + deadpool-postgres | 0 external crossbeam already in tree | Sub-10ms latency is guaranteed by design: zero allocation on the hot path, monomorphic dispatch, cache-line-isolated atomics, and a single-pointer CAS for connection handoff. ┌─────────────────┐ ┌─────────────────┐ │ TCP Client │ │ HTTP Client │ │ port 3000 │ │ port 3001 │ └────────┬────────┘ └────────┬────────┘ └─────────────┬─────────────┘ │ ┌────────┴────────┐ │ JSON-RPC 2.0 │ │ MCP Protocol │ └────────┬────────┘ │ ┌────────────┴────────────┐ │ Tool Dispatcher │ │ 135 tools │ └────────────┬────────────┘ │ ┌────────────┴────────────┐ │ Connection Pool │ │ lock-free, CAS-only │ │ Min: 5, Max: 20 │ └────────────┬────────────┘ │ ┌────────┴────────┐ │ PostgreSQL DB │ └─────────────────┘ Stateless HTTP — Each request is independent. Transaction state isolated per-connection. Lock-free pooling — crossbeam::ArrayQueue with CachePadded atomics. Zero mutex acquisitions. Input validation at the boundary — SQL capped at 10K chars, identifiers at 255, batch rows at 1K. SQL injection prevention via quote ident . PG version-aware queries — Verified against PG 16–18. Graceful fallbacks when views/columns differ across versions. Follows Semantic Versioning https://semver.org . The current line is 5.x , which targets MCP revision 2025-11-25 . The 5.0.0 release changed the tools/call result shape to be spec-compliant — see MIGRATION.md and the CHANGELOG /corporatepiyush/mcp-pg-rust/blob/main/CHANGELOG.md . | mcp-postgres | MCP revision default | Negotiates | |---|---|---| | 5.x | 2025-11-25 | 2025-06-18 , 2025-03-26 , 2024-11-05 | | ≤ 4.x | 2024-11-05 | — | Apache-2.0