Convert Markdown to Jira / Confluence Wiki Markup

Headings, bold, and a link in one ticket description

The everyday case: a ticket body with a heading, emphasis, and a doc link. Note bold collapses from two asterisks to one, and the link flips to the pipe form.

Markdown in:
## Summary
The **login** page 500s when the [SSO config](https://wiki/sso) is empty.

Jira out:
h2. Summary
The *login* page 500s when the [SSO config|https://wiki/sso] is empty.

Fenced code block with a language tag

Fenced code is stashed before the inline passes, so the asterisks and backticks inside the code are never touched. The fence language carries into the macro.

Markdown in:
```python
def f(x):
    return x ** 2  # not bold
```

Jira out:
{code:python}
def f(x):
    return x ** 2  # not bold
{code}

Nested unordered list

Two leading spaces equal one nesting level. The transform divides the indent length by two and repeats the bullet, so a two-space indent becomes a single extra asterisk.

Markdown in:
- Parent
  - Child
    - Grandchild

Jira out:
* Parent
** Child
*** Grandchild

Inline code, italic, and strikethrough together

Inline code becomes the {{…}} macro, italic becomes underscores, and strikethrough becomes single hyphens — all in one pass over the prose.

Markdown in:
Run `npm ci`, not ~~npm install~~. This is *important*.

Jira out:
Run {{npm ci}}, not -npm install-. This is _important_.

A GFM table that the tool leaves alone

There is no table-handling code in the transform. The table passes through verbatim and renders as monospaced plain text in Jira unless you rebuild it. This is the single most common surprise.

Markdown in:
| Env | Status |
| --- | ------ |
| prod | up |

Jira out (unchanged — needs manual rebuild):
| Env | Status |
| --- | ------ |
| prod | up |

Manual Jira form:
||Env||Status||
|prod|up|

The transform has no table logic. A | col | col | table passes through as literal pipe text and Jira shows it as plain monospaced lines, not a grid. Rebuild it in Jira's ||header||header|| plus |cell|cell| syntax by hand. This corrects a claim in older copy that pipe tables auto-convert — they do not.

An image ![alt](url) is not handled as an image. The [alt](url) portion matches the link rule and produces ![alt|url], which is neither Markdown nor Jira's !url! embed. After converting, replace these with !url! or upload the image as a Jira attachment and reference it.

If you paste the output into Jira's WYSIWYG editor, you will see literal h1., *, and {code} characters because the editor treats them as plain text. Switch to the text / wiki-markup input mode for that field before pasting, then toggle back to verify.

Jira treats _ as italic. A snippet of prose containing user_id or MAX_RETRIES (outside a code fence) can render the middle as italic in Jira. Wrap such tokens in inline code in your source so they become {{user_id}}, which Jira renders as monospace and ignores the underscores.

- [ ] do thing becomes * [ ] do thing because only the bullet marker is rewritten. Jira wiki markup has no native checkbox, so the [ ] / [x] text remains. Strip the brackets or use a Jira checklist app for interactive checkboxes.

The blockquote rule matches > text line by line. A lazy Markdown blockquote where only the first line has > will convert just that line to bq.; continuation lines stay as plain prose. Prefix every quoted line with > in the source for a clean conversion.

Because **x** to *x* and *x* to _x_ run as separate regex passes, deeply nested or adjacent emphasis like ***both*** may not produce the combination you expect. Review mixed-emphasis spans in the output and adjust by hand if needed.

Running with no input simply returns an empty string — there is nothing to convert. Paste your Markdown or select a file before clicking Run.

A large runbook or merged doc can exceed the Free cap of 1 MB / 500,000 characters. The character limit is checked independently of byte size, so a paste can hit it first. Split the document, or upgrade to Pro (10 MB / 5,000,000 chars) for larger inputs.