This MCP server provides read-only access to Iceberg tables through Apache Impala. It enables large language models (LLMs) to inspect database schemas and execute read-only SQL queries against Iceberg-backed data. The server exposes two core operations: execute_query(query: str), which runs any SQL query on Impala and returns the results as JSON, and get_schema(), which lists all tables available in the current database. It connects to Impala using environment-configured credentials and database settings (IMPALA_HOST, IMPALA_PORT, IMPALA_USER, IMPALA_PASSWORD, IMPALA_DATABASE). The transport layer is configurable via MCP_TRANSPORT, supporting stdio (default), http, and SSE, enabling local tools, web deployments, and existing web-based integrations. The repository provides deployment instructions for Claude Desktop and local setups, including GitHub-based and local directory installations. Examples in the repository showcase integration with AI frameworks like LangChain/LangGraph and the OpenAI SDK. This server is intended to help AI applications discover Iceberg schemas and fetch read-only results without modifying data.