Hello, @blankbro! I'm here to help you with any bugs, questions, or contributions!

To resolve the "chunk is not good" Markdown file parsing error in Dify version 0.6.11 when importing the README.md file from the mqtt-storm GitHub repository, follow these steps:

1. Check for Unclosed Tags: Ensure that all HTML tags within the Markdown file are properly closed.
2. Validate Markdown Syntax: Use a Markdown validator to check for any syntax errors.
3. Simplify Content: Temporarily remove complex elements like tables, images, and HTML to identify if any specific part is causing the issue.

Here is an example of a simplified and properly structured Markdown file:

# Project Title

![cover-image](https://example.com/image.png)

<p align="center">
  <a href="https://example.com">Link 1</a> ·
  <a href="https://example.com">Link 2</a> ·
  <a href="https://example.com">Link 3</a>
</p>

## Features

- Feature 1
- Feature 2
- Feature 3

## Usage

Instructions on how to use the project.

## Contributing

Guidelines for contributing to the project.

## License

Information about the project's license.

If the error persists, it may be necessary to debug further by isolating specific sections of the Markdown file to identify the problematic chunk ^[1].

Additionally, Dify uses the react-markdown library for Markdown parsing. Known issues with react-markdown version 8.0.6 include:

1. HTML Parsing: By default, react-markdown does not support parsing HTML within Markdown.
2. Plugins Compatibility: Some plugins or extensions might not be fully compatible.
3. Syntax Highlighting: Additional configuration or plugins like react-syntax-highlighter might be needed ^[2].

Dify also provides built-in tools for validating or sanitizing Markdown files before parsing them, such as replacing continuous spaces, newlines, and tabs, and deleting all URLs and email addresses ^[3].

For more detailed error messages, check the Logs section in the Dify application, accessible via the left-side navigation ^[4].

To continue talking to Dosu, mention @dosu.

Please use the custom separate identifier to do it. :)

Please use the custom separate identifier to do it. :)

@crazywoola

Is that it? But my problem still exists.

Yes, you need to choose which identifier you would like to use. For example, you could fill in ###

Yes, you need to choose which identifier you would like to use. For example, you could fill in ###

problem still exists.

Your doc is segmenting by ##, I mean for example ###, you should change it according to your needs.

@crazywoola hi, After my investigation, I found that it was caused by the commit #2528. The md file was processed and the pictures and links were automatically removed.