REST API

Web scraping for Dify AI applications

Add web tools to Dify agents and workflows.

Dify is an open-source LLM application development platform. Add webclaw as a custom tool via the Tools > Custom section, and your Dify agents can scrape, extract, and research the web at runtime.

Setup

Dify custom tool (OpenAPI-style)

openapi: 3.0.0
info:
  title: webclaw
  description: Web extraction API for LLM agents
  version: 1.0.0
servers:
  - url: https://api.webclaw.io
paths:
  /v1/scrape:
    post:
      summary: Scrape any URL to markdown
      operationId: scrape
      security:
        - bearerAuth: []
      requestBody:
        content:
          application/json:
            schema:
              type: object
              required: [url]
              properties:
                url:
                  type: string
                formats:
                  type: array
                  items:
                    type: string
      responses:
        '200':
          description: Success
components:
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer

Why webclaw for Dify

  • Registers as a standard OpenAPI custom tool
  • Works with Dify agents and workflows
  • Bot bypass for production Dify apps
  • LLM-optimized markdown works with Dify models

Common use cases

  • Dify agents with real-time web access
  • Chatbots that answer from live documentation
  • Knowledge workflows with fresh web data
  • RAG pipelines built in Dify

Frequently asked questions

How do I register webclaw as a Dify custom tool?

Go to Tools > Custom in your Dify workspace. Paste the OpenAPI spec above and configure your webclaw API key as the bearer token. The tool will be available to all agents in the workspace.

Can Dify workflow steps call webclaw?

Yes. Once registered as a custom tool, webclaw endpoints appear in workflow nodes as callable actions. You can scrape, extract, or search as part of any Dify flow.

Ready to connect?

Start extracting.

7-day Starter trial. Cancel anytime. Deploy in under a minute or self-host forever. Open source.