Markdown to reStructuredText

Yes, that is the main reason to use it rather than reformatting by hand. RST is stricter than Markdown about underlined section headers, directive indentation and blank lines, and pandoc's RST writer handles all of that. Headings become underlined sections, code fences become code blocks, and lists get the indentation RST needs.

It has no reStructuredText equivalent, so expect it to come out as a raw block or to disappear entirely. Read the output before you commit it. The same caution applies to anything else that is Markdown-flavoured rather than structural, like HTML entities or inline style attributes.

It is plain reStructuredText, not Sphinx-flavoured. Sphinx-only directives such as toctree, and roles such as ref and doc, are not invented by pandoc; you add those yourself. What pandoc gives you is a correct .rst body that Sphinx will happily parse once you drop it into your source tree.

Your file name with the extension replaced by .rst, so guide.md becomes guide.rst. The converted text is also shown on the page so you can read it before downloading, and copying takes the whole thing even when the preview is truncated.

Use reStructuredText to Markdown, the same pandoc engine in reverse. Round trips lose things in both directions: RST directives with no Markdown equivalent going one way, raw HTML going the other. Nothing is uploaded in either direction; pandoc runs as WebAssembly in this tab.