MCP 服务器

Flowise MCP Server

A Model Context Protocol server that enables LLM tools to create, manage, and run Flowise chatflows and agentflows programmatically.

README

Flowise MCP Server

A Model Context Protocol (MCP) server that provides programmatic integration with Flowise AI workflow platform. This enables LLM-based tools like Claude Code to create, manage, and run Flowise chatflows and agentflows.

Features

Run Predictions: Execute chatflows with questions, conversation history, file uploads, or lead capture
Manage Chatflows: Create, update, delete, and list chatflows programmatically
Node Discovery: List all available nodes and get detailed specifications for building flows
Full Flow Types: Supports CHATFLOW, AGENTFLOW, MULTIAGENT, and ASSISTANT types

Prerequisites

Node.js 20+
pnpm package manager
A running Flowise instance

Installation

# Clone the repository
git clone https://github.com/wksbx/flowise-mcp-server.git
cd flowise-mcp-server

# Install dependencies
pnpm install

# Build the project
pnpm build

Configuration

Copy the example environment file:

cp .env.example .env

Edit .env with your Flowise settings:

FLOWISE_BASE_URL=http://localhost:3000
FLOWISE_API_KEY=your-api-key-here

FLOWISE_BASE_URL: URL where your Flowise instance is running
FLOWISE_API_KEY: API key from Flowise (Settings > API Keys)

Usage

Running Directly

pnpm start

Running with Docker

# Build the Docker image
pnpm docker:build

# Run the container
pnpm docker:run

Configuring with MCP Clients

Add to your MCP client configuration (e.g., Claude Desktop, Claude Code):

Using Node directly:

{
  "mcpServers": {
    "flowise": {
      "command": "node",
      "args": ["/path/to/flowise-mcp-server/dist/index.js"],
      "env": {
        "FLOWISE_BASE_URL": "http://localhost:3000",
        "FLOWISE_API_KEY": "your-api-key-here"
      }
    }
  }
}

Using Docker:

{
  "mcpServers": {
    "flowise": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "--add-host=host.docker.internal:host-gateway",
        "--env-file", "/path/to/your/.env",
        "flowise-mcp-server"
      ]
    }
  }
}

Available Tools

Prediction Tools

Tool	Description
`create_prediction`	Run a chatflow with a question
`create_prediction_with_history`	Run with conversation history for context
`create_prediction_with_files`	Run with file attachments (images, documents)
`create_prediction_with_lead`	Run and capture lead email

Chatflow Management

Tool	Description
`list_chatflows`	List all available chatflows
`get_chatflow`	Get a specific chatflow's configuration
`create_chatflow`	Create a new chatflow
`update_chatflow`	Update an existing chatflow
`delete_chatflow`	Delete a chatflow (irreversible)

Node Discovery

Tool	Description
`list_nodes`	List all available node types
`get_nodes_by_category`	Get nodes filtered by category
`get_node`	Get detailed spec for a specific node type

Examples

Running a Chatflow

Use create_prediction with:
- chatflowId: "abc123"
- question: "What is the weather today?"

Creating a Simple Chatflow

1. Use get_node to fetch specs for needed nodes (e.g., "chatOpenAI", "llmChain")
2. Use create_chatflow with:
   - name: "My Chatflow"
   - flowData: { nodes: [...], edges: [...] }
   - type: "CHATFLOW"

Development

# Build TypeScript
pnpm build

# Run in development mode (build + run)
pnpm dev

Testing

The project includes comprehensive unit tests using Vitest.

# Run tests once
pnpm test

# Run tests in watch mode
pnpm test:watch

# Run tests with coverage report
pnpm test:coverage

Test Structure

src/
├── flowise-api.test.ts   # API client tests (8 tests)
└── handlers.test.ts      # Tool handler tests (26 tests)

Project Structure

flowise-mcp-server/
├── src/
│   ├── index.ts           # MCP server entry point
│   ├── flowise-api.ts     # Flowise API client
│   ├── handlers.ts        # Tool handler functions
│   └── *.test.ts          # Unit tests
├── dist/                  # Compiled JavaScript (generated)
├── package.json
├── tsconfig.json
├── vitest.config.ts       # Test configuration
├── Dockerfile
├── .env.example           # Environment template
└── mcp-config.example.json

Troubleshooting

Connection Issues

Ensure Flowise is running and accessible at the configured URL
When using Docker, use host.docker.internal to connect to Flowise on the host machine
Verify your API key is correct in Flowise settings