wa:confluence-publish

You can publish Markdown documentation to Confluence by scanning files under docs/, computing content hashes, and creating or updating pages only when changes are detected. It requires MCP connectivity and CONFLUENCE_ENABLED gating before executing operations.

Incremental sync for Confluence page updates works by computing local content hashes and comparing them against stored metadata. It creates missing pages and updates existing ones only when needed, then records Confluence page IDs and last published metadata in local frontmatter after successful operations.

Yes, you can use frontmatter to control Confluence page creation and labeling. The Skill reads document frontmatter, including type and status, to decide whether to create or update a page, and applies labels based on the doc type using the Confluence API’s native labels parameter.

Yes, Confluence publishing requires MCP preflight checks. The Skill mandates MCP connectivity via core/ops:mcp-preflight and enforces CONFLUENCE_ENABLED gating before attempting to publish any local documentation to ensure operational safety.

The best way to test Confluence publishing without making live changes is to use the dry-run mode. This allows you to target specific paths or patterns and preview the success, skip, and failure report without actually creating or updating Confluence pages.

Your Confluence pages may not be updating because the content hashes have not changed, triggering an incremental sync skip. You can use the force mode to override this behavior and update pages regardless of computed content hashes.