notion-formatter_skill

Overview

This skill formats Markdown for smooth Notion import and paste. It converts standard Markdown, applies Notion-friendly toggles, blockquotes, code-block labels, and pipe-table syntax. It annotates features that require manual recreation in Notion and flags common import pitfalls. The goal is predictable, editable content after pasting into Notion.

How this skill works

The formatter scans Markdown and enforces conventions Notion expects: toggles use "> ", blockquotes use '" ', fenced code blocks include language identifiers, and tables use pipe syntax. It replaces or annotates unsupported items (equations, complex embeds) with clear [NOTION: ...] notes so you can recreate them manually. It also validates image URLs, local path warnings, and basic structural checks before output.

When to use it

• Preparing AI-generated responses for direct paste into Notion
• Converting existing .md documentation into Notion-compatible Markdown
• Creating Notion-friendly export from technical notes with code and tables
• Preparing mixed-content pages with images, tables, or collapsible sections
• Splitting large documents into importable chunks

Best practices

• Keep document structure shallow; avoid deep nesting of lists or toggles
• Label fenced code blocks with the language (```python) to preserve syntax blocks
• Use > (greater-than + space) for toggles and " (quote + space) for blockquotes consistently
• Annotate anything Notion cannot recreate (equations, complex embeds) with [NOTION: ...]
• Verify image links are public URLs and replace local paths before importing

Example use cases

• Format a chat response that contains code, a table, and notes for import into a Notion page
• Convert a project's README sections into toggles so each section becomes collapsible in Notion
• Prepare API docs with language-labeled code samples and pipe-style tables for clean import
• Split a long guide into smaller chunks and mark manual polish items for final editing in Notion
• Annotate equations and advanced embeds so reviewers know which blocks need manual recreation