Markdown-Gfm
Markdown-Gfm是一款productivity方向的AI技能,核心价值是Markdown formatting for GitHub-flavored markdown (GFM) files,可用于解决开发者在productivity领域的实际问题,帮助用户提升效率、自动化重复任务或优化工作流。
Markdown formatting for GitHub-flavored markdown (GFM) files
mkdir -p ./skills/markdown-gfm && curl -sfL https://raw.githubusercontent.com/github/awesome-copilot/main/skills/markdown-gfm/SKILL.md -o ./skills/markdown-gfm/SKILL.md Run in terminal / PowerShell. Requires curl (Unix) or PowerShell 5+ (Windows).
Skill Content
# GitHub Flavored Markdown (GFM)
Apply these rules per the [GFM spec](https://github.github.com/gfm/) when writing or reviewing `.md` files. GFM is a strict superset of CommonMark. GFM spec for reference only. Do not download GFM Spec.
Preliminaries
- A line ends at a newline (`U+000A`), carriage return (`U+000D`), or end of file. A blank line contains only spaces or tabs.
- Tabs behave as 4-space tab stops for block structure but are not expanded in content.
- Replace `U+0000` with the replacement character `U+FFFD`.
Leaf Blocks
- **Thematic breaks**: 3+ matching `-`, `_`, or `*` characters on a line with 0–3 spaces indent. No other characters on the line. Can interrupt a paragraph.
- **ATX headings**: 1–6 `#` characters followed by a space or end of line. Optional closing `#` sequence (preceded by a space). 0–3 spaces indent allowed.
- **Setext headings**: Text underlined with `=` (level 1) or `-` (level 2). Cannot interrupt a paragraph — blank line required after a preceding paragraph.
- **Indented code blocks**: Lines indented 4+ spaces. Cannot interrupt a paragraph. Content is literal text, not parsed as Markdown.
- **Fenced code blocks**: Open with 3+ backticks or tildes (do not mix). Closing fence must use same character with at least the same count. Specify language identifier after the opening fence. Content is literal text.
- **HTML blocks**: Seven types defined by start/end tag conditions. Types 1–6 can interrupt paragraphs; type 7 cannot. Content is passed through as raw HTML.
- Type 1: `<script>`, `<pre>`, or `<style>` (case-insensitive) — ends at matching closing tag.
- Type 2: `<!--` comment — ends at `-->`.
- Type 3: `<?` processing instruction — ends at `?>`.
- Type 4: `<!` + uppercase letter (e.g., `<!DOCTYPE>`) — ends at `>`.
- Type 5: `<![CDATA[` — ends at `]]>`.
- Type 6: Block-level HTML tags (`<div>`, `<table>`, `<p>`, `<h1>`–`<h6>`, `<ul>`, `<ol>`, `<section>`, etc.) — ends at a blank line.
- Type 7: Any other complete open or closing tag on its own line — ends at a blank line. Cannot interrupt a paragraph.
- **Link reference definitions**: `[label]: destination "title"`. Case-insensitive label matching. First definition wins for duplicate labels. Cannot interrupt a paragraph.
- **Paragraphs**: Consecutive non-blank lines not interpretable as other block constructs. Leading spaces up to 3 are stripped.
- **Blank lines**: Ignored between blocks; determine whether a list is tight or loose.
- **Tables** *(extension)*: Header row, delimiter row (`---`, `:---:`, `---:`), zero or more data rows. Delimit cells with `|`. Escape literal pipe as `\|`. Header and delimiter must have matching column count. Broken at first blank line or other block-level structure.
Container Blocks
- **Block quotes**: Lines prefixed with `>` (optionally followed by a space). Lazy continuation allowed for paragraph text only. A blank line separates consecutive block quotes.
- **List items**: Bullet markers (`-`, `+`, `*`) or ordered markers (1–9 digits + `.` or `)`). Content column determined by marker width + spaces to first non-whitespace. Sublists must be indented to the content column. An ordered list interrupting a paragraph must start with `1`.
- **Task list items** *(extension)*: `- [ ]` (unchecked) or `- [x]` (checked) at the start of a list item paragraph. Space between `-` and `[` is required. May be nested.
- **Lists**: Sequence of same-type list items. Changing bullet character or ordered delimiter starts a new list. A list is loose if any item is separated by a blank line.
Inlines
- **Backslash escapes**: `\` before any ASCII punctuation character renders the literal character. Not recognized in code spans, code blocks, or autolinks.
- **Entity and numeric character references**: `&`, `{`, `{` — valid HTML5 entities. Not recognized in code spans or code blocks. Cannot replace structural characters.
- **Code spans**: Backtick-delimited inline code. Line endings convert to s
🎯 Best For
- Claude users
- GitHub Copilot users
- Knowledge workers
- Remote teams
- Professionals
💡 Use Cases
- Using Markdown-Gfm in daily workflow
- Automating repetitive productivity tasks
📖 How to Use This Skill
- 1
Install the Skill
Copy the install command from the Terminal tab and run it. The SKILL.md file downloads to your local skills directory.
- 2
Load into Your AI Assistant
Open Claude or GitHub Copilot and reference the skill. Paste the SKILL.md content or use the system prompt tab.
- 3
Apply Markdown-Gfm to Your Work
Provide context for your task — paste source material, describe your audience, or share existing work to guide the AI.
- 4
Review and Refine
Edit the AI output for accuracy, tone, and completeness. Add human insight where the AI lacks context.
❓ Frequently Asked Questions
How do I install Markdown-Gfm?
Copy the install command from the Terminal tab and run it. The skill downloads to ./skills/markdown-gfm/SKILL.md, ready to use.
Can I customize this skill for my team?
Absolutely. Edit the SKILL.md file to add team-specific instructions, examples, or workflows.
⚠️ Common Mistakes to Avoid
Not reading the full skill
Skills contain important context and edge cases beyond the quick start.