ClawHub

Clean Text Formatter

Other

Removes Markdown formatting and excess whitespace from text to produce clean, publication-ready plain text with normalized punctuation and spacing.

Install

openclaw skills install clean-text-formatter

Clean Text Formatter

Name

clean-text-formatter

Description

Removes Markdown formatting characters and excessive whitespace from AI-generated text. Produces clean, publication-ready plain text.

Capabilities

1. Markdown Stripping

Strips the following while preserving content:

  • Headers: #, ##, ### (removes markers, keeps text)
  • Bold/italic markers: **, __, *, _ (removes markers, keeps text)
  • Links: [text](url) → keeps text
  • Images: ![alt](url) → keeps alt
  • Blockquotes: > lines (removed)
  • Code blocks: backticks removed, code content preserved
  • List markers: -, *, 1. (removed, list content preserved)
  • Horizontal rules: ---, *** (removed)
  • Footnotes: [^1] (removed)

2. Whitespace Cleaning Around Numbers

Fixes common AI spacing issues:

  • 100 % → 100%
  • ¥ 100 → ¥100
  • USD 50 → USD 50 (currency symbols kept with numbers)
  • Numbers with leading/trailing spaces trimmed
  • Thousand separators , preserved: 1,000,000

3. Punctuation Spacing

  • Removes spaces before punctuation: ， 。 ！ ？
  • Removes spaces inside parentheses: （ 文本 ） → （文本）
  • Normalizes multiple spaces to single space
  • Trim leading/trailing whitespace from each line

4. Unicode Normalization

  • Normalizes quotes: "text" → "text", 'text' → 'text'
  • Normalizes dashes: — – → -

Triggers

  • User pastes text and asks to "clean" or "remove Markdown"
  • User shares AI-generated content and asks to "format for publication"
  • Any content containing visible Markdown syntax

Workflow

  1. Receive raw text input (paste, file upload, or document content)
  2. Detect content type and Markdown density
  3. Apply cleaning rules in order: a. Structural (headers, lists, blockquotes) b. Inline formatting (links, images, emphasis) c. Whitespace normalization d. Punctuation spacing fix
  4. Return clean text + optional export

Input

  • Plain text paste
  • .txt, .md, .docx, .html files
  • Direct document content

Output

  • Clean plain text (ready to copy/paste)
  • Optional: .txt file export

Limitations

  • Does not preserve Markdown intended for re-use (strips permanently)
  • Complex nested formatting may require manual review
  • LaTeX math expressions: $$...$$ and $...$ are preserved

Dependencies

  • Python 3.8+ (for regex processing)
  • No external packages required