md2html.sh

#!/usr/bin/env bash

# Download HTML converted from provided Markdown, using GitHub API v3
##
md2html(){
  if [[ $# -ne 2 ]]; then
    echo "ERROR.\nSYNTAX: Markdown_To_HTML <markdown-filepath> <dest-html-filepath>"
    return
  fi
  unset _markdown_filepath
  unset _html_filepath
  unset _github_json_filepath
  unset _markdown_for_github

  _markdown_filepath=$1
  _html_filepath=$2
  _github_json_filepath="${_markdown_filepath}.json"

  sed -i 's/$/\\n/g' $_markdown_filepath
  sed -i 's/"/\\"/g' $_markdown_filepath

  echo "{ \"text\" : \"" > $_github_json_filepath
  cat $_markdown_filepath >> $_github_json_filepath
  echo "\" }" >> $_github_json_filepath

  cat $_github_json_filepath | curl -sLk -X POST -d@- https://api.github.com/markdown > $_html_filepath
  rm $_github_json_filepath
  echo "Successful conversion of ${_markdown_filepath} to ${_html_filepath}."
}

GitHub has a "raw" version of this endpoint just to make this sort of thing easier. So this whole thing could be replaced with the following alias:

$ alias md='curl https://api.github.com/markdown/raw -s -XPOST -H Content-Type:text/plain -d @-'

Usage:

$ echo '**Hello, World!**' | md
<p><strong>Hello, World!</strong></p>

Of course, to use it more like this script, you could read and write files:

$ md < test.md > test.html