For some HTML mails, it is sufficient to use a mailcap file in combination with lynx or w3m:

  # ~/.mutt/mailcap
  text/html; w3m -I %{charset} -T text/html; copiousoutput;

  # ~/.muttrc
  alternative_order text/plain text/enriched text/html  # save html for last
  auto_view text/html

If you want the full HTML experience, you can use this script [1], which uses python (there also is an older bash version) to convert the mail into standalone html to be viewed inside a browser.

  # ~/.muttrc
  macro index,pager M "<enter-command>unset resolve<enter><clear-flag>N<enter-command>set resolve<enter><pipe-message>viewhtmlmail\n" "View HTML in browser"

1: https://github.com/akkana/scripts/blob/master/viewhtmlmail

The reverse should work for simple HTML.

For complex HTML to plain text try https://www.google.com/search?q=lynx+convert+html+text .. most text browsers can do a reasonable job of it.

Using lynx (or elinks, in my case) to view HTML email in mutt is the historical solution. I think the desire here is for a reversible transformation. E.g. I can receive an HTML email, view it as Markdown, reply using Markdown, and then send an HTML email in reply.

Unfortunately, I think foreign email clients would mangle the HTML just enough to stop this from working. You'd have to start implementing a tag soup type parser to coax incoming mail into something intelligible.

We're almost at that stage with HTML already.

I'm thinking that a multi-pass parser:

1. Is this sane HTML5 markup? I'll eat that.

2. Is this a borked-but-manageable variant? Strip the cruft.

3. Is this Beyond All Hope? Just dump it as ASCII. It'll be blockquoted in reply.

HTML is no longer a peresistent, stable, or universal interchange format.

Well you can use pandoc too or w3m.