sqlsense and llm-sql-safety-executor-mcp
About sqlsense
raj8github/sqlsense
Safe, audited SQL for AI agents via MCP.
Implements query-level guardrails (blocking dangerous operations, auto-limiting SELECTs, column masking) that execute before reaching the database, with full audit logging of agent activity. Runs as an MCP server compatible with Claude Desktop and Claude Code, supporting PostgreSQL, MySQL, SQL Server, Snowflake, BigQuery, DuckDB, and SQLite through a unified interface with configurable readonly mode and per-table/column access controls.
About llm-sql-safety-executor-mcp
Firstmeridian/llm-sql-safety-executor-mcp
A secure database access gateway for AI Agents: Empowering LLM (Agents) with database access capabilities.
Implements an MCP (Model Context Protocol) server that enforces SQL safety through parsing validation, allowlists, and result truncation, while exposing tools like `query()`, `list_tables()`, and `describe_table()` to guide LLM decision-making via the ReAct pattern. Supports MySQL and SQLite with SQLAlchemy connection pooling, and integrates with AutoGen multi-agent frameworks and MCP clients like Claude Desktop and VS Code Copilot.
Scores updated daily from GitHub, PyPI, and npm data. How scores work