Automate Confluence Page Creation and Updates

This skill automates Confluence operations - page creation and updates, content search with Confluence Query Language (CQL), space management, label tagging, and page hierarchy navigation - through Composio's Confluence toolkit accessed via Rube MCP. Setup requires adding https://rube.app/mcp as an MCP server (no API keys needed), confirming RUBE_SEARCH_TOOLS responds, calling RUBE_MANAGE_CONNECTIONS with toolkit confluence, and completing Confluence OAuth if the connection isn't already ACTIVE. Creating and updating pages follows a tool sequence of CONFLUENCE_GET_SPACES to find the target space, optionally CONFLUENCE_SEARCH_CONTENT to avoid duplicate titles, CONFLUENCE_GET_PAGE_BY_ID to fetch the current version before any update, then CONFLUENCE_CREATE_PAGE or CONFLUENCE_UPDATE_PAGE, and optionally CONFLUENCE_ADD_CONTENT_LABEL - noting that Confluence enforces unique titles per space, updates require version.number set to current version + 1, content must be Confluence storage format (XHTML) rather than Markdown, and CREATE_PAGE nests content under body.storage.value while UPDATE_PAGE uses body.value plus body.representation. Content search uses CONFLUENCE_SEARCH_CONTENT for keyword ranking or CONFLUENCE_CQL_SEARCH for true full-text search with CQL fields (text, title, label, space, type, creator, lastModified, ancestor), operators (=, ~, >, IN, etc.), and functions like currentUser() and now("-7d") - noting that SEARCH_CONTENT only client-side filters up to 300 pages rather than doing true full-text search, that body expansions in CQL search reduce the max result limit to 25-50, and that recent pages may not appear immediately due to indexing lag. Space management covers listing, creating, and inspecting spaces via CONFLUENCE_GET_SPACES/GET_SPACE_BY_ID/CREATE_SPACE/GET_SPACE_PROPERTIES/GET_SPACE_CONTENTS/GET_LABELS_FOR_SPACE, with space keys restricted to alphanumeric characters and GET_SPACE_BY_ID requiring the numeric space ID rather than the key. Page hierarchy and label navigation covers GET_CHILD_PAGES (direct children only, cursor-paginated, requiring recursion for a full tree), GET_PAGE_ANCESTORS, GET_LABELS_FOR_PAGE/GET_LABELS_FOR_SPACE_CONTENT, ADD_CONTENT_LABEL, and GET_PAGE_VERSIONS for edit-history auditing. Common patterns cover resolving human-readable names to IDs before any operation, the two pagination styles Confluence uses (offset-based start/limit for most endpoints, cursor-based via _links.next for child pages), and content-formatting rules (XHTML storage format with <p>/<table>/<ac:structured-macro> elements, never Markdown). Known pitfalls to watch for: numeric space/page IDs versus short space keys, HTTP 429 rate limits requiring throttling to roughly 2 requests/second with backoff and jitter, invalid XHTML failing page creation, and version conflicts from concurrent edits requiring an immediate re-fetch of the current version before any update.

Use this skill when creating or updating Confluence pages, searching Confluence content with keyword or CQL search, managing spaces, navigating page hierarchies, or managing labels - all through Rube MCP's Confluence toolkit. It requires an active Rube MCP connection with the confluence toolkit authenticated via OAuth; it is not usable without that connection in place. Notably, CONFLUENCE_CREATE_PAGE accepts a space key directly, so a space doesn't always need to be resolved to its numeric ID first, and it is not the right tool for removing a page (that's the separate CONFLUENCE_DELETE_PAGE call).

Input is a Confluence operation request (create/update a page, search content, manage a space, navigate hierarchy, or manage labels) plus any required parameters (space key/ID, page title, CQL query, etc.). Output is the corresponding Confluence state change or query result - a created/updated page, matched search results, space metadata, or a page's children/ancestors/labels/version history - executed through the resolved Rube MCP tool calls.

Built on Rube MCP (https://rube.app/mcp) and Composio's Confluence toolkit, covering CONFLUENCE_GET_SPACES, CONFLUENCE_CREATE_PAGE, CONFLUENCE_UPDATE_PAGE, CONFLUENCE_SEARCH_CONTENT, CONFLUENCE_CQL_SEARCH, CONFLUENCE_GET_CHILD_PAGES, CONFLUENCE_ADD_CONTENT_LABEL, and related space/page/label/version tools.

title ~ "meeting" AND lastModified > now("-7d") ORDER BY lastModified DESC

Teams automating Confluence documentation workflows - page creation/updates, content search, space management, and hierarchy/label navigation - through Rube MCP without managing Confluence API keys directly.