dbhub and postgres-mcp

These are direct competitors: both provide PostgreSQL database access via MCP, but Bytebase's tool supports multiple database engines with higher adoption, while the Postgres-specific alternative targets token efficiency through a single consolidated interface.

dbhub
80
Verified
postgres-mcp
45
Emerging
Maintenance 23/25
Adoption 20/25
Maturity 18/25
Community 19/25
Maintenance 13/25
Adoption 4/25
Maturity 18/25
Community 10/25
Stars: 2,287
Forks: 187
Downloads: 71,034
Commits (30d): 32
Language: TypeScript
License: MIT
Stars: 6
Forks: 1
Downloads:
Commits (30d): 0
Language: TypeScript
License: MIT
No risk flags
No risk flags

About dbhub

bytebase/dbhub

Zero-dependency, token-efficient database MCP server for Postgres, MySQL, SQL Server, MariaDB, SQLite.

Implements just two core MCP tools—`execute_sql` and `search_objects`—with optional custom parameterized operations defined in TOML config, minimizing token overhead for AI context windows. Supports stdio, HTTP, and SSE transports for integration with Claude Desktop, VS Code, Cursor, and other MCP clients, plus includes a built-in web workbench for direct database interaction. Enforces safety guardrails including read-only mode, row limits, and query timeouts while enabling secure multi-database connections through SSH tunneling and SSL/TLS.

About postgres-mcp

neverinfamous/postgres-mcp

PostgreSQL MCP Server: Secure Administration & Observability Featuring Code Mode—One Tool Replacing All Specialized 232 Tools for up to 90% Token Savings. Includes Connection Pooling, HTTP/SSE, OAuth 2.1, Deterministic Error Handling and Full Support for 8 Extensions (citext, HypoPG, ltree, pgcrypto, pg_cron, pg_stat_kcache, pgvector & PostGIS).

Implements a JavaScript sandbox (Code Mode) that executes multi-step database operations locally within a single tool invocation, enabling AI agents to chain 232 PostgreSQL capabilities without round-trip communication overhead. Provides 20 observability resources for real-time schema, performance, and connection metrics, plus 19 guided prompts for query optimization and extension management. Targets AI assistants and agents via MCP with dual HTTP/SSE transport, schema introspection, migration tracking with SHA-256 deduplication, and granular OAuth 2.1 scopes (`read`, `write`, `admin`, `db:*`, `table:*:*`).

Scores updated daily from GitHub, PyPI, and npm data. How scores work