dify-workflow

Generate importable Dify 0.6.0 workflow DSL files from natural language descriptions.

152|27|Updated Apr 10, 2026
One-click install
npx skills add https://github.com/LingyiChen-AI/workflow-skill --skill dify-workflow-lingyichen-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dify-workflow
Source: https://github.com/LingyiChen-AI/workflow-skill/tree/main/skills/dify-workflow
Command: npx skills add https://github.com/LingyiChen-AI/workflow-skill --skill dify-workflow-lingyichen-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It turns a natural-language description of a business process into a fully specified Dify workflow DSL (YAML by default, JSON on request), so you can import and run the workflow without manually designing nodes, edges, schemas, and layout.

Core Features & Use Cases

  • Natural-language-to-DSL generation: Scaffolds complete Dify graph definitions from your requirement (app name, mode, node flow, and configuration).
  • Schema-aware node assembly: Selects the right node types (Start/End/LLM/Tool/HTTP/IF-ELSE/Iteration/Knowledge Retrieval, etc.) and fills required fields according to Dify 0.6.0 DSL structure.
  • Correct edges + routing logic: Generates valid edge IDs, handles branching conventions (if-else and question classifier), and computes node positions on the canvas.
  • Chatflow vs workflow mode support: Produces advanced-chat graphs using Start/Answer nodes or workflow graphs using Start/End nodes, matching the required variable and system variable rules.
  • Practical deployment patterns: Works well for common workflows such as RAG chatbots, tool-using agents, multi-branch assistants, and parameter-extraction flows.

Quick Start

Ask the skill to generate a Dify workflow DSL for your use case, then import the returned .dify.yml into Dify using the UI “Import DSL”.

Frequently Asked Questions about dify-workflow

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a Dify workflow DSL from natural language?

Generate a Dify workflow DSL from natural language by providing a process description, and the system outputs a valid Dify 0.6.0 YAML or JSON file with correct node schemas and edges. You can then import this file directly via the Dify UI.

Can I create a multi-node Dify graph with LLM calls and conditionals using plain text?

Yes, you can create a multi-node Dify graph with LLM calls, conditionals, iterations, and HTTP requests using plain text. The system assembles the correct node schemas, wires edges, and computes deterministic layout rules for the canvas.

Does Dify DSL generation support both advanced-chat and workflow modes?

Dify DSL generation supports both advanced-chat and workflow modes. It produces advanced-chat graphs using Start and Answer nodes, and workflow graphs using Start and End nodes, matching required variable and system variable rules.

What is the best way to build a RAG chatbot workflow for Dify without manual node design?

The best way to build a RAG chatbot workflow for Dify without manual node design is to describe the flow in natural language. The system scaffolds complete Dify graph definitions, including knowledge retrieval and LLM nodes, into an importable DSL file.

How do I import a generated Dify workflow DSL file into the platform?

To import a generated Dify workflow DSL file, use the "Import DSL" function in the Dify UI. The system outputs a valid .dify.yml file containing the complete graph definition, which you can upload and run immediately.

What are the limitations of generating Dify workflows from text descriptions?

Limitations of generating Dify workflows from text descriptions include strict adherence to the Dify 0.6.0 DSL structure and deterministic layout rules. The generated file requires correct variable references and edge wiring, meaning incomplete text descriptions may produce incomplete node configurations.