How to Publish Multi-File GitHub Markdown to Confluence Cloud Without Broken Links

Why Single-File Markdown Importers Break Repository Documentation

When engineering teams copy or import Markdown files one by one into Confluence Cloud, relative file paths like ../architecture/security-model.md or ./assets/diagram.png either point to 404 GitHub paths or break completely. Furthermore, unguarded scripts overwrite manual hotfixes made by product managers or SREs directly in Confluence.

<!-- Broken single-file import -->
<a href="../setup.md#install">Setup Guide</a>

<!-- DocBrancher Pass-2 Confluence Storage XHTML -->
<ac:link ac:anchor="install">
  <ri:page ri:content-title="Setup &amp; Provisioning Guide" ri:space-key="ENG" />
  <ac:plain-text-link-body><![CDATA[Setup Guide]]></ac:plain-text-link-body>
</ac:link>

Two-Pass Tree Resolution & Read-Only Destination Preflight

Before writing a single byte to Confluence Cloud, DocBrancher runs a read-only preflight check against the destination space. The preflight plan classifies every selected Markdown file as create, update, skip (when the SHA-256 content hash matches), or conflict (when the remote Confluence page version advanced past the last synced version).

Version-Checked Human-Edit Conflict Protection

If a colleague edits a Confluence page directly in the browser (advancing the remote page from v3 to v4), DocBrancher pauses publication for that page and presents a side-by-side comparison between the remote Confluence Storage XHTML and the proposed GitHub commit.

Frequently Asked Questions