AgentSkillsCN

confluence-reader

将 Confluence Cloud 页面以 Markdown 格式读取，浏览页面树，下载图片附件，并将本地内容与 Confluence 页面进行差异对比。当用户提及 Confluence、维基页面、Confluence 文档，或需要将本地文档与 Confluence 进行比对时，可使用此功能。

SKILL.md
--- frontmatter
name: confluence-reader
description: Read Confluence Cloud pages as markdown, browse page trees, download image attachments, and diff local content against Confluence pages. Use when the user mentions Confluence, wiki pages, Confluence documentation, or needs to compare local docs with Confluence.
license: MIT
compatibility: Requires Python 3.9+ and the requests library (auto-installed if missing). Environment variables CONFLUENCE_TOKEN, CONFLUENCE_EMAIL, and one of CONFLUENCE_CLOUD_ID or CONFLUENCE_BASE_URL must be set.
metadata:
  author: satiyap
  version: "0.2.3"
  source: https://github.com/satiyap/confluence-reader-mcp

Confluence Reader

Read and interact with Confluence Cloud pages by running scripts/confluence.py.

Prerequisites

These environment variables must be set:

  • •CONFLUENCE_TOKEN — Scoped API token
  • •CONFLUENCE_EMAIL — Email tied to the Atlassian account
  • •CONFLUENCE_CLOUD_ID or CONFLUENCE_BASE_URL — one is required

If any are missing, prompt the user to set them before proceeding.

Commands

All commands are run via:

bash
python scripts/confluence.py <command> [args...]

Fetch a page as markdown

bash
python scripts/confluence.py fetch-page "<confluence-page-url>"

Returns the page content as markdown with child pages listed at the bottom.

List child pages

bash
python scripts/confluence.py list-children "<confluence-page-url>"

Returns titles and IDs of direct child pages without fetching content.

Download an image attachment

bash
python scripts/confluence.py fetch-image "<confluence-page-url>" "<filename>" "<destination-dir>"

Downloads the named attachment and saves it to the destination directory.

Compare local content with a Confluence page

bash
python scripts/confluence.py compare "<confluence-page-url>" "<local-markdown-or-filepath>"

The last argument can be a file path or a raw markdown string. Returns a JSON object with additions, deletions, totalChanges, and the full unified diff.

Supported URL formats

  • •/wiki/spaces/SPACEKEY/pages/123456789/Page+Title
  • •/wiki/pages/viewpage.action?pageId=123456789

Workflow: Recursively reading a page tree

  1. •Run fetch-page on the root URL.
  2. •Note the child pages listed at the bottom.
  3. •Run fetch-page for each child.
  4. •Repeat until all desired pages are fetched.

Workflow: Syncing local docs

  1. •Read the local markdown file.
  2. •Run compare with the Confluence URL and the local file path.
  3. •If totalChanges > 0, present the diff to the user.

Edge cases

  • •Empty pages return an empty markdown body — normal for container pages.
  • •If an attachment filename doesn't match, the script prints available attachments.
  • •401/403 errors mean invalid or under-permissioned credentials.