confluence-automation

Confluence Automation - Rube MCP Agent Skill

Skill Overview

Automate Confluence operations via Rube MCP to create and update pages, perform CQL content searches, manage spaces, and batch-handle labels — connect without an API key.

Applicable Scenarios

1. Bulk Document Creation and Maintenance

When you need to create technical documents, API references, or user manuals in bulk, you can automatically generate Confluence pages with formatted content, support XHTML storage format, and automatically add labels for categorization.

2. Knowledge Base Content Search and Management

Use CQL (Confluence Query Language) for advanced searches to quickly locate content by space, time range, or labels. Supports full-text search and structured queries, suitable for routine maintenance of large knowledge bases.

3. Space and Hierarchy Management

Automate space creation, page hierarchy traversal, and child page management; recursively retrieve full page tree structures. Suitable for project documentation initialization, space migration, or content reorganization.

Core Features

1. Page Creation and Version Management

Supports creating new pages in a specified space, submitting content in XHTML storage format. Automatically handles version numbers when updating pages to avoid concurrent edit conflicts. The page’s position in the hierarchy can be controlled via the parentId parameter.

2. CQL Advanced Search

Provides true full-text search capabilities, supporting multi-dimensional filtering by title, body, labels, creator, modification time, etc. CQL supports complex query conditions such as text ~ "keyword" AND lastModified > now("-7d"), and allows custom sorting and result expansion.

3. Space and Label Management

List all available spaces, with filtering by type and status. Can create new spaces (note that the space key may only contain alphanumeric characters). Supports adding and querying labels at both page and space levels for easy categorization and retrieval.

Frequently Asked Questions

How do I connect Confluence to Rube MCP?

First add https://rube.app/mcp as an MCP server — no API key required. Then call RUBE_MANAGE_CONNECTIONS and select the confluence toolkit, complete authorization via the returned OAuth link, and ensure the status is ACTIVE before use.

What’s the difference between CQL search and regular search?

CONFLUENCE_SEARCH_CONTENT is client-side filtering, fetching up to 300 pages and then filtering; CONFLUENCE_CQL_SEARCH is a server-side, true full-text search that supports complex condition combinations and indexed queries. CQL is better suited for large spaces and precise search needs.

How do I avoid Confluence update version conflicts?

Before updating a page, you must call GET_PAGE_BY_ID to get the current version number, then set version.number to current version + 1. It is recommended to fetch version information immediately before updating to avoid conflicts from concurrent edits.