risingwave

RisingWave MCP Server exposes a rich set of tools that let you query and manage a RisingWave streaming database via natural language. Once the MCP server is running, you can interact with it through supported clients (e.g., VS Code Copilot or Claude Desktop) by issuing requests that map to a variety of tools across categories like Query, Explain, Schema, DDL, DML, and data source/sink management. Typical workflows include listing tables, inspecting table schemas, creating or altering materialized views, querying data, and monitoring streaming jobs or backfills. The server is built on FastMCP and uses a high-performance STDIO transport, enabling responsive AI-assisted interactions with the underlying RisingWave instance. You can start with simple questions such as “List my tables” or “Show me the backfill progress for streaming jobs,” then progressively perform complex operations like creating materialized views or altering source configurations through the MCP tools.

The Available Tools cover: Query tools (run_select_query, table_row_count, get_table_stats), Explain tools (explain_query, explain_analyze, explain_distsql), Schema tools (show_tables, describe_table, show_create_table, get_relation_info, list_schemas, etc.), DDL tools (create_materialized_view, drop_table, add_table_column, execute_ddl_statement, etc.), DML tools (insert_single_row, update_rows, delete_rows), Source tools (list_sources, describe_source, refresh_source_schema, etc.), Sink tools (describe_sink, rename_sink, drop_sink, etc.), Connection tools, Secret tools, Index tools, and UDF tools. This makes it suitable for both ad-hoc data exploration and programmatic management via your AI assistant.

Prerequisites:

• Python 3.8+ installed on your system
• Git installed
• Internet access to fetch dependencies

Step 1: Clone the MCP server repository

git clone https://github.com/risingwavelabs/risingwave-mcp.git
cd risingwave-mcp

Step 2: Create a Python environment (recommended)

python -m venv venv
# On Windows
venv\Scripts\activate
# On macOS/Linux
source venv/bin/activate

Step 3: Install Python dependencies

pip install -r requirements.txt

Step 4: Run the MCP server locally for development

python src/main.py

Step 5: (Optional) Configure integration with your IDE or Claude Desktop per the README guidance.

Notes and tips:

• Ensure RisingWave is running and accessible before starting the MCP server (local or cloud deployment).
• The server communicates via STDIO transport; make sure the client tooling you use supports STDIO-based MCP communication.
• If you encounter connection issues, verify environment variables or connection strings used by the client configuration (e.g., host, port, user, password, and database).
• For VS Code Copilot and Claude Desktop, follow the integration sections in the README to wire up server commands and environment variables (RISINGWAVE_HOST, RISINGWAVE_PORT, etc.).
• The MCP server supports a broad suite of tools; start with list-related queries (e.g., show_tables, describe_table) to familiarize yourself with the responses before attempting DDL or data manipulation tasks.