md-to-lexical

name	md-to-lexical
description	Convert Shiroi Markdown to Haklex Lexical JSON. Use when the user wants to convert a markdown file or string to rich text editor JSON format.
user_invocable	true

Markdown to Lexical JSON Converter

Converts Shiroi-flavored Markdown to Haklex's Lexical SerializedEditorState JSON.

Usage

Run the conversion script:

npx tsx scripts/md-to-lexical/index.ts <input.md> > output.json

Or pipe from stdin:

cat article.md | npx tsx scripts/md-to-lexical/index.ts - > output.json

Supported Syntax

| Markdown | Lexical Node | | ---------------------------------------------- | ---------------------------------- | ---- | --------- | --- | ----------- | | # heading | HeadingNode | | paragraph | ParagraphNode | | > quote | QuoteNode | | > [!NOTE/TIP/WARNING/IMPORTANT/CAUTION] | AlertQuoteNode | | - list / 1. list | ListNode | | --- | HorizontalRuleNode | | ```lang ``` | CodeBlockNode | | ```mermaid ``` | MermaidNode | | ![alt](src "caption") | ImageNode | | ![alt](video.mp4) | VideoNode (auto-detect) | | <video src="..."> | VideoNode | | | table | | TableNode | | $eq$ | KaTeXInlineNode | | $$eq$$ | KaTeXBlockNode | | **bold**, *italic*, ~~strike~~ | text format flags | | ++underline++, ==highlight== | text format flags | | ^superscript^, ~subscript~ | text format flags | | `code` | text format flag | | | | text | | | SpoilerNode | | [name]{GH@handle} | MentionNode | | [^id] / [^id]: ... | FootnoteNode / FootnoteSectionNode | | <ruby>base<rt>reading</rt></ruby> | RubyNode | | <tag>text</tag> | TagNode | | [text](url) | LinkNode | | standalone URL paragraph | LinkCardNode | | :::gallery/carousel | GalleryNode | | :::warn/info/error/banner | BannerNode | | :::grid{cols=N,gap=M} | GridContainerNode | | :::masonry | GalleryNode (masonry) | | <details><summary>...</summary>...</details> | DetailsNode | | <Tabs><tab label="...">code</tab>...</Tabs> | CodeSnippetNode | | <excalidraw>snapshot</excalidraw> | ExcalidrawNode | | <nested-doc>content</nested-doc> | NestedDocNode |

Workflow

1. User provides a markdown file path or pastes markdown content
2. Run the script to generate JSON
3. Output can be saved to a file or used directly