Markdown ZIP export

Create an export request for one publication, inspect its state, and download a ZIP archive. The exported archive contains unarchived posts, authors, categories, and tags; it does not include other publication resources.

Authentication

All routes require a bearer token and the listed export capability in the selected publication. Export capabilities are available to publication owners and trusted internal principals.

Routes

Route	Method	Capability	Required input	Optional input and behavior	Success result
/export/start	POST	export.start	Publication selector	format	Staged export.
/export/status	POST	export.status	export_id	Matching publication selector	export.
/export/download	POST	export.download	export_id	Matching publication selector	ZIP attachment.
/export/cancel	POST	export.cancel	export_id	Matching publication selector	Cancelled export.

Start and download

format defaults to and currently only accepts markdown_zip.

curl -sS https://api.blog.service.m7.org/v1/export/start \
  -H 'Authorization: Bearer ACCESS_TOKEN' \
  -H 'Content-Type: application/json' \
  -d '{"publication_slug":"engineering-notes","format":"markdown_zip"}'

{
  "started": true,
  "export": {
    "export_id": "EXPORT_UUID",
    "publication_id": "PUBLICATION_UUID",
    "format": "markdown_zip",
    "state": "ready",
    "selection": { "content": "posts", "include_archived": false },
    "archive_filename": "engineering-notes-export-20260819-120000.zip",
    "summary": { "post_count": 12, "author_count": 2, "category_count": 3, "tag_count": 9 }
  }
}

Use export_id with status, download, and cancel. Download succeeds only while the export state is ready; it returns an application/zip attachment, not the JSON envelope. Use the Content-Disposition filename when saving the response. Cancel changes a ready request to cancelled and does not change publication content.

Archive layout

manifest.json
authors.json
categories.json
tags.json
posts/
  release-notes.md

Each post uses importer-compatible front matter. The archive reflects current unarchived data at download time, rather than a permanent snapshot taken at start.