mcp-doris
An MCP server for Apache Doris & VeloDB
claude mcp add --transport stdio morningman-mcp-doris python -m mcp_doris.mcp_server \ --env DORIS_HOST="host of Doris" \ --env DORIS_PORT="port of Doris" \ --env DORIS_USER="Doris username" \ --env DORIS_PASSWORD="Doris password"
How to use
This MCP server integrates Apache Doris into the Model Context Protocol ecosystem, exposing Doris-specific capabilities to describe and interact with your Doris data sources in the MCP framework. The server is driven by a Python-based MCP implementation and relies on standard Doris connection environment variables to establish access. Typical workflows involve preparing a Doris connection, inspecting or synchronizing metadata, and then using the MCP UI or API to explore or test queries against Doris through the MCP abstraction.
To run locally, configure the Doris connection environment variables (host, port, user, password) and start the server via the provided Python module. Once the server is running, you can use the MCP Inspector via the CLI to sync and then access the web UI at the local host where the inspector serves (as documented in development steps). The available tooling emphasizes synchronization (to pull metadata into MCP), and then browsing or validating the Doris models within the MCP environment.
How to install
Prerequisites:
- Python 3.8+ (recommended)
- A Doris instance accessible from your development environment
- Basic tooling for MCP development (git, Python virtual environment)
Step-by-step:
-
Clone the repository containing the MCP server: git clone <repository-url> cd mcp-doris
-
Create and activate a Python virtual environment: python -m venv .venv source .venv/bin/activate # on Unix/macOS .venv\Scripts\activate # on Windows
-
Install required Python dependencies (if a requirements.txt exists): pip install -r requirements.txt
or install specific dependencies for the Doris MCP server as documented
-
Configure Doris connection via environment variables (example): export DORIS_HOST=localhost export DORIS_PORT=8030 export DORIS_USER=doris_user export DORIS_PASSWORD=secret
-
Run the MCP server module: python -m mcp_doris.mcp_server
-
If you use the MCP Inspector during development, you can follow the project’s development workflow (sync, activate virtualenv, set PYTHONPATH, then start the dev server): uv sync source .venv/bin/activate export PYTHONPATH=/path/to/mcp-doris:$PYTHONPATH env DORIS_HOST=$DORIS_HOST DORIS_PORT=$DORIS_PORT DORIS_USER=$DORIS_USER DORIS_PASSWORD=$DORIS_PASSWORD mcp dev mcp_doris/mcp_server.py
-
Access the MCP web interface at http://localhost:5173 (as per the development notes).
-
For publishing or building workflows, you can follow the project’s publish steps (if provided): uv build, uv publish
Additional notes
Tips and common issues:
- Ensure Doris is reachable from your environment and that network ACLs allow the connection.
- The DORIS_* environment variables are required to establish the connection; avoid hard-coding credentials in scripts.
- When using uv-based development commands, ensure you are in the correct virtual environment and PYTHONPATH includes the MCP server path.
- If the MCP UI does not load, verify that the server module path (mcp_doris.mcp_server) is correct and that the Doris client dependencies are installed.
- Check Doris connection details (host, port, user, password) for correctness and permissions.
- The server may expose a local web interface during development; use the provided URL (http://localhost:5173) to inspect metadata and test MCP commands.
- If you modify the MCP server, re-run the server to pick up changes or use a development workflow described in the repository docs.
Related MCP Servers
Wax
Sub-Millisecond RAG on Apple Silicon. No Server. No API. One File. Pure Swift
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
mcp-yfinance
Real-time stock API with Python, MCP server example, yfinance stock analysis dashboard
mcp_autogen_sse_stdio
This repository demonstrates how to use AutoGen to integrate local and remote MCP (Model Context Protocol) servers. It showcases a local math tool (math_server.py) using Stdio and a remote Apify tool (RAG Web Browser Actor) via SSE for tasks like arithmetic and web browsing.
cloudwatch-logs
MCP server from serkanh/cloudwatch-logs-mcp
servicenow-api
ServiceNow MCP Server and API Wrapper