I have a Github Markdown document that contains HTML tags - mostly to do with special characters (e.g. ℋ) and stuff to place pictures where I want in the page. The thing is, pandoc seems to ignore the HTML tags. Is this a limitation intrinsic to pandoc, or is there any way to get pandoc to process such tags and produce the right output? 

The pandoc invocation that I am currently using for converting my Github Markdown documents to PDF is

 $ pandoc --resource-path=/home/abc/Repos.wiki -t html5 --pdf-engine=wkhtmltopdf --metadata pagetitle="MyDoc.md" --css github.css -o MyDoc.pdf

The default invocation pandoc MyDoc.md -o MyDoc.pdf is not dealing with images properly (in that it sometimes rearranges surrounding paragraphs the wrong way) and it seems to be unable to deal with expressions like _A__{_m_}, in that the _and directives seem to be ignored.