iFlytek Workflow

This a simple implementation of an MCP server using iFlytek. It enables calling iFlytek workflows through MCP tools.

python

27

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "iflytek-ifly-workflow-mcp-server": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/iflytek/ifly-workflow-mcp-server",
        "ifly_workflow_mcp_server"
      ],
      "env": {
        "CONFIG_PATH": "$CONFIG_PATH"
      }
    }
  }
}

This MCP server lets you call iFlytek workflows from your LLM applications by exposing MCP tools that invoke the iFlytek workflow runtime. It enables automated workflow orchestration with configurable nodes, execution modes, and multi-model support, so you can schedule, branch, loop, and stream results as part of your AI-enabled processes.

How to use

Prepare your workflow configuration and connect your MCP client to start invoking iFlytek workflows. You create a local config that points to your workflow data, then register the MCP server in your client configuration. The server supports starting from a user input (Start Node) and delivering output (End Node), with execution managed automatically according to your predefined flow.

To use the server from an MCP client, provide a workflow plan via a config file and bind the MCP server in your client with the appropriate authentication. You can publish workflows and obtain the necessary API bindings to call the workflows from your agent or app. The key steps are: create a bot, publish and debug your workflow, publish as API, configure the application binding, and retrieve the workflow ID and authentication information.

How to install

Prerequisites: ensure you have access to the MCP runtime that provides the uvx-based command flow shown below.

Add the MCP server configuration to your client config file. The following JSON snippet demonstrates how to register the iFlytek workflow MCP server under the name ifly-workflow-mcp-server.

{
  "mcpServers": {
    "ifly-workflow-mcp-server": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/iflytek/ifly-workflow-mcp-server",
        "ifly_workflow_mcp_server"
      ],
      "env": {
        "CONFIG_PATH": "$CONFIG_PATH"
      }
    }
  }
}

Additional configuration details

Place the CONFIG_PATH to point to your local workflow config file, which contains the flow_id and API key information for authenticating with the cloud or runtime you are using.