Skip to content
Get packs

Guide · Open source

How to use Awesome-Dify-Workflow: clone, import and adapt Dify DSL files

Updated · written with AI from the project's own documentation and checked against our catalogue data

Awesome-Dify-Workflow is an MIT-licensed collection of Dify workflow definition files (DSL .yml files) shared by the author for personal use and learning. It is aimed at people who already run Dify and want working examples to study or adapt. After this guide you will know which prerequisites to check, how to get the repository, and how to import a workflow into your own Dify account.

Difysvcvit/Awesome-Dify-Workflow · ★ 10.8k · MIT

Quick start

  1. 1Check your Dify version. Confirm that your Dify installation is version 0.13.0 or later, and use the latest version if you need Agent nodes.
  2. 2Clone the repository. Run git clone https://github.com/svcvit/Awesome-Dify-Workflow to get the DSL files.
  3. 3Pick a workflow. Read the descriptions under the update-date headings and choose the entry closest to your goal.
  4. 4Copy the DSL file URL. Find the matching file in the DSL folder and copy its URL, as the README instructs.
  5. 5Import into Dify. Import the file into your own Dify account.
  6. 6Set up any companion pieces. If the entry names a plugin, project or key, install or configure it before running the workflow.
  7. 7Review and adapt. Check each node, replace placeholders and test the workflow with your own inputs.

Check the prerequisites before you start

The README asks you to import these files into Dify 0.13.0 or later. It recommends the latest Dify version because Agent nodes were added after version 1.0.

You need your own Dify setup. The collection contains workflow files, not a Dify installation. The README suggests either following the official guide to deploy Dify locally with Docker, or using the hosted Dify cloud service, where free users can create up to 5 workflows.

  • Dify 0.13.0 or later (latest recommended for Agent nodes)
  • A Dify account, either self-hosted with Docker or on the hosted cloud
  • Git, to clone the repository

Get the repository with git clone

Clone the repository to your computer. The README gives this command, and the DSL files are kept in the DSL folder.

git clone https://github.com/svcvit/Awesome-Dify-Workflow

Choose a workflow from the DSL list

The README groups workflows by update date. Each entry has a file name, a short description and a source link. Read the description to find a workflow that fits your goal, such as a chatbot template, a tool workflow, or an example of MCP tool calls or HTML artifact rendering.

Some entries are not standalone DSL files. For example, the sanic-web entry links to a separate project, and the Artifact.yml entry needs a companion Dify plugin. Read the entry before you import it.

  • MCP-amap.yml shows MCP tool calls through the MCP Agent strategy and needs a free AMap API key that you add yourself
  • Artifact.yml needs the dify-plugin-artifacts plugin and an extension setup, as described in its entry
  • 图文知识库 is a knowledge base example that includes a markdown file

Import the DSL file into your Dify account

The README says to copy the file URL from the DSL folder and import it into your own Dify account. Then adapt the workflow to your needs, for example by replacing placeholder keys or changing model settings.

The workflows are described as typically built from basic inputs, condition judgement, a variable aggregator and output nodes. They can be published as a tool and embedded in a chatbot flow.

  • Open the DSL file in the DSL folder and copy its URL
  • Import it into your Dify account, then review each node before running it
  • Replace any placeholder keys or sample settings the README mentions

Fix common errors using the README notes

The README includes short answers to frequent problems. These fixes are specific to self-hosted Docker setups and are written in Chinese in the source.

For sandbox package errors, the README says to add the dependency to /docker/volumes/sandbox/dependencies/python-requirements.txt and restart the sandbox. If pandas, numpy above 2.0, matplotlib or scikit-learn still fail with errors such as operation not permitted, the author points to a separate project, dify-sandbox-py, where those dependencies are said to have been tested. For string-length limits between nodes, the README suggests changing CODE_MAX_STRING_LENGTH and TEMPLATE_TRANSFORM_MAX_LENGTH in .env and restarting the containers.

CODE_MAX_STRING_LENGTH: 1000000 TEMPLATE_TRANSFORM_MAX_LENGTH: 1000000

Explore the plugin examples

Since Dify 1.0 introduced plugins, the README also points to examples of plugin code. The repository dify_plugin_collection stores plugin packages downloaded from the official Dify marketplace. The author also published four plugins with open source code: google_translate (a simple Tool), tod_agent (an Agent strategy) and Artifacts (an Extension that renders HTML).

These are useful for reading how plugin file structures work. The README warns that the Agent strategy is harder to build and needs both product thinking and development skill.

  • dify-plugin-google_translate: a small Tool to study the plugin file structure
  • dify-plugin-tod_agent: an Agent strategy example
  • dify-plugin-artifacts: an Extension example for HTML rendering

What people use it for

  • Study how workflow nodes, variables and forms are connected in a working example
  • Start from a template for a chatbot or tool workflow and change it to fit your task
  • Try an MCP tool call through the MCP Agent strategy with the MCP-amap.yml example
  • Test HTML rendering of LLM output with the Artifact.yml workflow and the Artifacts plugin

Frequently asked questions

Is Awesome-Dify-Workflow free to use?

The repository is licensed under MIT. The README says that basically all of the workflows can be used for free, and that more are still being collected. Some entries depend on third-party services, which you may need to sign up for separately.

What do I need before importing these workflows?

You need Dify 0.13.0 or later, with the latest version recommended for Agent nodes. You also need your own Dify account, since the files are imported into it.

How do I install Awesome-Dify-Workflow?

The README gives one command: git clone https://github.com/svcvit/Awesome-Dify-Workflow. After cloning, you copy the URL of the DSL file you want and import it into Dify. The repository itself is not a program you install.

Why do my pandas or numpy packages fail in the sandbox?

The README says the sandbox can fail with errors such as operation not permitted when using pandas, numpy above 2.0, matplotlib or scikit-learn. The author suggests the separate dify-sandbox-py project, where these dependencies are said to have been tested.

Are there alternatives for learning Dify?

The README references dify101.com as a learning site for self-study. It also lists related projects such as sanic-web, a database Q&A project that uses Dify as its service layer, and dify-schedule, which the README links for scheduled tasks.

Can I use these workflows in production?

The README says the workflows are shared for personal use and learning, and it does not claim they are production-grade or fully tested. Review and test any workflow carefully before relying on it.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Awesome-Dify-Workflow suits developers and students who already run Dify and want real workflow files to study or adapt. It is not a good fit for people who have not set up Dify yet, or for anyone who needs production-grade, fully tested workflows.

Install options & downloads