ms82119/NiFiMCP
If you are the rightful owner of NiFiMCP and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
NiFi MCP provides a chat interface for interacting with Apache NiFi using natural language queries, leveraging LLM and MCP tools.
NiFi MCP (Model Context Protocol)
This repository contains the NiFi MCP project. It provides a chat interface allowing users to interact with Apache NiFi instances using natural language queries. The system uses a Large Language Model (LLM) integrated with custom tools (MCP Tools) that can communicate with the NiFi API to retrieve information, document flows, and create and perform actions on NiFi components.
It has been tested with Nifi versions 1.23 and 1.28, but could work on other versions assuming the Nifi REST Api stays consistent. While it is quite functional, you will find that the type of LLM model you use will have a big effect on how well it follows your instructions. I have found o4-mini and gpt-4.1 to be good.
It's ability to read and document existing flows is very good. It's ability to create new and modify existing flows is OK, but often takes several iterations and some user help. My future plans are to refine the tools to improve this aspect. I may also introduce some tools to help it debug flows itself.
Example Conversations
- - See how the system creates a complete NiFi flow from requirements
- - Watch the system diagnose and fix issues in an existing flow
Recent Updates
For the latest updates and release notes, see the GitHub Releases page.
Roadmap
My future plan is to work on improving the tool's accuracy and efficiency (reducing the number of tokens needed to acheive the goal). To do this I will focus on introducing guided workflows which can provide more granular control over the steps and quality, which may also lead to the ability to use less powerful and cheaper LLM's while improving the results. I feel creating new flows and documenting large flows will particularly benefit from this.
After this I will consider looking into support for more models. I will say that the Gemini support has been harder than expected due to its protobuf usage, and has slowed my progress along the roadmap, so for this reason I am going to hold off adding other models for now.
Setup Instructions
Note: These instructions should also be followed after pulling a new version as there may be new package requirements.
To set up the development environment and install all dependencies, follow these steps:
-
Clone the Repository:
git clone git@github.com:ms82119/NiFiMCP.git cd NiFiMCP -
Set Up a Virtual Environment: It's recommended to use a virtual environment to manage dependencies. You can create one using
venv:python3 -m venv .venv -
Activate the Virtual Environment:
- On macOS and Linux:
source .venv/bin/activate - On Windows:
.venv\Scripts\activate
- On macOS and Linux:
-
Install Dependencies: Use
uvto install dependencies based onpyproject.tomlanduv.lock:uv sync -
Update the config.yaml file with your Nifi details and LLM API keys Use the config.example.yaml as your guide for format and structure
-
Run the MCP Server: Start the MCP server with:
uvicorn nifi_mcp_server.server:app --reload --port 8000 -
Run the Streamlit Client: Start the Streamlit client with:
python -m streamlit run nifi_chat_ui/app.py
Usage Tips
For detailed usage information, tips, and UI features, see the .
Running Automated Tests
For comprehensive testing information and examples, see the .
