Bazhuayu Dify Plugin

This Dify tool plugin wraps the Bazhuayu Streamable HTTP MCP server and exposes Bazhuayu cloud scraping tools inside Dify.

It supports the core scraping workflow:

1. search_templates - find a cloud-runnable scraping template
2. execute_task - validate parameters, create a cloud task, and start it
3. export_data - check progress and export structured data

It also exposes task management and promotion helpers:

4. search_tasks - search existing cloud tasks
5. start_or_stop_task - start or stop an existing task
6. redeem_coupon_code - redeem a Bazhuayu coupon or promotion code

How it works

The plugin does not reimplement Bazhuayu scraping logic. It forwards Dify tool calls to the existing Bazhuayu MCP server:

Dify Tool Plugin -> Bazhuayu MCP Server -> Bazhuayu Client API

Default MCP endpoint:

https://mcp.bazhuayu.com/

You can override it in provider credentials or with:

BAZHUAYU_MCP_SERVER_URL=https://mcp.bazhuayu.com/

Authentication

The plugin supports API key and OAuth credential modes.

API Key

Configure:

api_key
mcp_server_url

The plugin sends:

X-API-Key: <api_key>

This reuses the API key path already implemented by the Bazhuayu MCP server.

OAuth

Click Add OAuth in Dify to authorize your Bazhuayu account.

The plugin uses the Bazhuayu OAuth client:

client_id: DifyMCP
client_secret: *
authorization_url: https://identity.bazhuayu.com/connect/authorize
token_url: https://identity.bazhuayu.com/connect/token
scope: openid profile offline_access

Dify handles user authorization and token refresh. The plugin then calls the MCP server with:

Authorization: Bearer <access_token>

This reuses the Bearer-token path already implemented by the Bazhuayu MCP server.

Tools

search_templates

Search cloud-runnable Bazhuayu templates.

Typical input:

{
  "keyword": "amazon",
  "limit": 8
}

Use recommendedTemplate.templateName from the response when calling execute_task.

execute_task

Validate parameters, create a cloud task, and start it.

Typical validation input:

{
  "templateName": "amazon-product-scraper",
  "validateOnly": true,
  "parameters": "{\"SearchKeyword\":[\"iphone\"]}"
}

Typical execution input:

{
  "templateName": "amazon-product-scraper",
  "parameters": "{\"SearchKeyword\":[\"iphone\"]}"
}

parameters is a JSON object string. Use the inputSchema[].field keys returned by search_templates or validateOnly.

export_data

Check task progress and export data.

Typical input:

{
  "taskId": "your-task-id",
  "exportFileType": "JSON",
  "previewRows": 5
}

Supported export types:

JSON, CSV, EXCEL, HTML, XML

search_tasks

Search existing Bazhuayu cloud tasks.

Typical input:

{
  "keyword": "amazon",
  "page": 1,
  "size": 10
}

You can also filter by status:

{
  "status": "Running"
}

Supported status values:

Running, Stopped, Completed, Failed

start_or_stop_task

Start or stop an existing Bazhuayu cloud task.

Typical input:

{
  "taskId": "your-task-id",
  "action": "start"
}

Supported actions:

start, stop

redeem_coupon_code

Redeem a Bazhuayu coupon or promotion code.

Typical input:

{
  "code": "RESOURCE_TMPL_01"
}

Development

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env
python -m main

Package

Install the Dify plugin CLI, then run:

dify plugin package .

The generated .difypkg can be submitted to langgenius/dify-plugins.