Confluence Cloud

io.github.aaronsbPublished by GitHub account aaronsbv0.6.0Updated Oct 24, 2026

MCP server for Confluence Cloud — page editing, navigation, and content management

Overview

AI-generated overview

A local MCP server that lets an assistant read, search, edit, and organize Confluence Cloud pages, spaces, attachments, and page links.

What you can do
Get, create, update, delete, move, copy, or pull pages for editing, including session-based block edits that sync only changed sections instead of rewriting whole pages. Search with CQL, full-text, labels, or contributors, and list spaces or view space details. Upload, download, list, or delete page attachments. Traverse page hierarchy and discover backlinks, forward links, and related pages. Batch several operations into one queued call with result references and error strategies.
When to use it
Useful when an assistant should work inside an existing Confluence Cloud wiki: drafting or patching documentation, finding pages by label or contributor, reorganizing page trees, or managing attachments without manual copy-paste.
Requirements
Runs locally via npx from the npm package, so Node.js is needed. Requires three environment variables: CONFLUENCE_API_TOKEN (an Atlassian API token), CONFLUENCE_EMAIL (the Atlassian account email), and CONFLUENCE_HOST (the Confluence instance URL). Network access to the Confluence instance and Atlassian services is required.
Before you install
The API token grants access to your Confluence account, so store it carefully and scope it to what you need. The server can create, update, delete, move, and copy pages and delete attachments, so assistant actions can change or remove real content. Page content and metadata are sent to the configured Confluence instance and Atlassian services.

Installation

In SourceWeft

  1. Open Confluence Cloud in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

Confluence Cloud MCP Server

A Model Context Protocol server for interacting with Confluence Cloud. Structured page editing with session-based change tracking, native macro support, and graph-based navigation.

Install

Claude Desktop (one-click)

Download confluence-cloud-mcp.mcpb and open it — Claude Desktop will prompt for your Confluence credentials.

Claude Code

bash
claude mcp add confluence-cloud -e CONFLUENCE_API_TOKEN=your-token -e CONFLUENCE_EMAIL=your-email -e CONFLUENCE_HOST=https://your-team.atlassian.net -- npx -y @aaronsb/confluence-cloud-mcp

Manual (any MCP client)

json
{  "mcpServers": {    "confluence-cloud": {      "command": "npx",      "args": ["-y", "@aaronsb/confluence-cloud-mcp"],      "env": {        "CONFLUENCE_API_TOKEN": "your-api-token",        "CONFLUENCE_EMAIL": "your-email",        "CONFLUENCE_HOST": "https://your-team.atlassian.net"      }    }  }}

Credentials

Generate an API token at Atlassian Account Settings.

Tools

ToolDescription
manage_confluence_pageGet, create, update, delete, move, copy, or pull pages for editing
edit_confluence_contentStructural block editing within a tracked session — patch sections, append, replace, find/replace, sync
manage_confluence_spaceList spaces, get space details, or manage space configuration
search_confluenceSearch using CQL, full-text, labels, or contributors
manage_confluence_mediaUpload, download, list, or delete page attachments
navigate_confluenceTraverse page hierarchy, discover backlinks (via GraphQL), forward links, and related pages
queue_confluence_operationsBatch multiple operations with result references ($0.pageId) and error strategies

Each tool accepts an operation parameter (except queue_confluence_operations which takes an operations array). Per-tool documentation is available as MCP resources at confluence://tools/{tool_name}/documentation.

Key Features

Session-based editing — Pull a page into a tracked session, make surgical edits to individual blocks (sections, paragraphs, macros, tables), then sync only what changed. No full-page rewrites.

Native macro support — Status badges, info/warning/error panels, expand blocks, and table of contents render as readable :::directive syntax. The server handles ADF serialization with correct native node types.

GraphQL navigation — Backlinks and forward links use the Atlassian GraphQL gateway's link graph for accurate, fast relationship discovery. Falls back to REST when GraphQL is unavailable.

Rendering facades — Every response is token-efficient markdown with context-aware next-step hints. No raw JSON.

MCP Resources

ResourceDescription
confluence://macrosAvailable macro registry with parameter schemas and usage examples

Architecture

See docs/architecture/INDEX.md for the 8 ADRs covering the five-layer architecture, hybrid client, content model, session editing, macro handling, navigation, and rendering facades.

License

MIT License

Source: README.md at commit 821b467

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

2
  1. v0.6.0LatestSep 23, 2026
  2. v0.5.2Sep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts