marksmith 0.5.0

A Markdown toolbox — convert, template, and publish Markdown documents

marksmith

A Markdown toolbox — write docs in Markdown, ship them as polished DOCX or publish them to Confluence.

Installation

pip install marksmith

Optional extras for template support:

pip install marksmith[template]

Optional extras for Confluence publishing:

pip install marksmith[confluence]

---

Quick start

# Explicit output path
marksmith to-docx my-doc.md output.docx

# Output path taken from docx-path in the file's front-matter
marksmith to-docx my-doc.md

# Recursively convert all .md files in a directory tree that have docx-path set
marksmith to-docx --directory docs/

# Or via python -m
python -m marksmith to-docx my-doc.md output.docx

---

Markdown front-matter

You can add a YAML front-matter block at the top of your Markdown file. The metadata is written to the DOCX core properties (title, author, etc.) and is also available as Jinja2 template variables when using --template.

---
title:          My Document
version:        1.0
author:         Fred Bloggs
date:           2026-03-16
classification: Internal
docx-path:      '%MY_DOCS%\Reports\My Document.docx'
---

# My Document

Content goes here...

docx-path

The optional docx-path key sets the default output path for marksmith to-docx. If you omit the output argument on the command line, the path from front-matter is used. Environment variables are expanded automatically (%VAR% on Windows, $VAR on Linux/macOS).

Windows paths: Always use single quotes around the value in YAML. In double-quoted YAML strings, backslashes are escape characters and will corrupt the path.

# ✅ Correct
docx-path: '%MY_DOCS%\Reports\myfile.docx'

# ❌ Wrong — \R and \m etc. are YAML escape sequences
docx-path: "%MY_DOCS%\Reports\myfile.docx"

---

Supported Markdown elements

Element	Status
Headings H1 – H6	✅
Bold / italic / inline code	✅
Fenced and indented code blocks	✅
Unordered lists (nested)	✅
Ordered lists (nested)	✅
Block-quotes	✅
Tables	✅
Thematic breaks (horizontal rules)	✅
Strikethrough	✅
Links (text rendered, no hyperlink)	⚠️
Images	⚠️ placeholder text only
Inline HTML	➖ ignored

---

Template support

Keep your content in plain Markdown while producing brand-consistent DOCX output from a corporate template.

pip install marksmith[template]
marksmith to-docx my-doc.md output.docx --template company-template.docx

MARKSMITH_TEMPLATE environment variable

If you always use the same template, set MARKSMITH_TEMPLATE once in your environment and marksmith will use it automatically — no --template argument needed. An explicit --template argument always takes priority.

# Windows (set permanently via System Properties → Environment Variables)
$env:MARKSMITH_TEMPLATE = "C:\Users\You\Templates\company-template.docx"

# Then just run — template is applied automatically
marksmith to-docx my-doc.md

How it works

1. Create a .docx template in Word with Jinja2-style placeholders for metadata sourced from your Markdown front-matter:

   {{ title }}        {{ version }}      {{ author }}
   {{ date }}         {{ classification }}
   

2. Add a {{p marksmith_content }} paragraph alone on its own line at the exact point where the Markdown body should be inserted:

   {{p marksmith_content }}
   

   Important: Use {{p ... }} (with the p modifier), not {{ ... }}. The p modifier inserts a paragraph-level sub-document rather than plain text, so all headings, lists, tables, and code blocks are preserved.

3. Run the conversion — marksmith will:

   • Render all front-matter metadata into the Jinja2 placeholders.
   • Convert the Markdown body to native DOCX content.
   • Insert the body at {{p marksmith_content }}.
   • Save the merged document.

Style inheritance

Content is inserted as a sub-document, so heading and paragraph styles are matched by name against those in your template. If your template defines Heading 1 with a custom font and colour, the converted content will pick it up automatically.

---

Publishing to Confluence

Publish Markdown files directly to a Confluence space.

pip install marksmith[confluence]

Credentials

Create a .env file with your Confluence URL and a Personal Access Token. marksmith checks the following locations in order:

1. .env in the current working directory
2. .env in your home directory (~/.env on Linux/macOS, %USERPROFILE%\.env on Windows)

CONFLUENCE_URL=https://your-confluence.example.com
CONFLUENCE_TOKEN=your-personal-access-token

Front-matter fields

Each file must include a YAML front-matter block that tells marksmith where to publish the page:

---
title: My Page Title
space: MYSPACE
parent_title: My Parent Page   # resolved by title; or use parent_id: 12345
labels:
  - my-label
page-properties:
  description: A short description of this page
  created_by: Fred Bloggs
  version: "1.0"
  document_status: DRAFT       # DRAFT (yellow) | PUBLISHED (green) | other (red)
---

# My Page Title

Content goes here...

Field	Required	Description
title	Yes	Confluence page title
space	Yes	Confluence space key
parent_title	One of	Title of the parent page (resolved at publish time)
parent_id	One of	Numeric Confluence page ID of the parent
labels	No	List of labels to apply
page-properties	No	Renders a Page Properties macro at the top of the page

If the page already exists it is updated; otherwise it is created.

Commands

# Publish a single file
marksmith to-confluence --upload-file path/to/my-doc.md

# Publish all .md files in a directory
marksmith to-confluence --upload-directory path/to/docs/

# Publish a tree of repositories (each must have a confluence.yml at its root)
marksmith to-confluence --process-repo-tree path/to/repos/

Repository tree mode — confluence.yml

When using --process-repo-tree, each subdirectory that contains a confluence.yml is processed. The file defines the target space, parent page, and index page title for that repository:

repo-docs-path: docs/               # relative path to the docs directory
conflu-directory-title: My Service  # title of the index page created in Confluence
conflu-parent-id: "12345"           # Confluence page ID to create the index under
conflu-space-id: MYSPACE            # Confluence space key

All four fields are mandatory.

Additional features

Image attachments — local images referenced in Markdown are automatically uploaded as attachments to the published page.

Confluence link replacements — annotate a Markdown link with a <!-- replace_link(<page_id>) --> comment to rewrite it as a Confluence internal link at publish time:

[See related page](./other.md) <!-- replace_link(98765) -->

---

Roadmap

Action	Description
to-docx	Markdown → DOCX ✔️
to-docx --template	Merge into branded DOCX template ✔️
to-docx (docx-path)	Output path from front-matter docx-path key ✔️
to-docx --directory	Recursively convert all files with docx-path set ✔️
to-confluence	Publish Markdown to Confluence ✔️
lint	Validate Markdown style and structure
toc	Generate / update table of contents
diff	Show structural diff between two Markdown files
Images	Embed local images via run.add_picture()
Hyperlinks	Full OOXML hyperlink support

---

Development

git clone https://github.com/tkdpython/marksmith
cd marksmith
pip install -e .[dev,template]

# Run tests
pytest

# Lint
ruff check .

---

Releasing

1. Bump __version__ in marksmith/__init__.py.
2. Commit and push.
3. Create a GitHub Release with a tag matching the version (e.g. v0.2.0).
4. The publish workflow fires automatically and publishes to PyPI via OIDC Trusted Publisher — no API tokens needed.

---

License

MIT — see LICENSE.