cn

name: cn description: Use the cn CLI to interact with Confluence. Search pages, read content, create/update/delete pages, manage labels, move pages, check sync status, and more. Use when user wants to work with Confluence pages via cn CLI.

cn - Confluence CLI Skill

Use the cn CLI to interact with Confluence spaces and pages.

Scope

Only use cn commands. Do not reach for kb, MCP tools, or other search tools — if the user wanted those they would ask for them explicitly.

Config

Credentials stored at ~/.cn/config.json:

{
  "confluenceUrl": "https://company.atlassian.net",
  "email": "user@company.com",
  "apiToken": "..."
}

Run cn setup to configure interactively.

Commands Reference

Search

cn search 'type=page AND text~"query"'                              # Full-text search
cn search 'type=page AND space=ENG AND text~"query"'               # Narrow to a space
cn search 'type=page AND lastModified >= "2026-01-01"'             # Date filter
cn search 'type=page AND label=draft AND space=DOCS'               # Label filter
cn search 'type=page AND text~"query"' --limit 20                  # More results (default: 10)
cn search 'type=page AND text~"query"' --xml                       # XML output for parsing

The query argument is passed directly to the Confluence CQL search API. Use full CQL syntax for filtering by space, label, type, date, author, and more.

Browse & Explore

cn spaces                            # List spaces (first 25)
cn spaces --limit 50                 # More spaces
cn spaces --page 2                   # Next page of results
cn spaces --page 2 --limit 10        # Custom page size
cn spaces --xml                      # XML output for parsing

cn tree                              # Show page hierarchy (current dir)
cn tree ENG                          # Show tree for a specific space
cn tree --depth 2                    # Limit to 2 levels deep
cn tree --xml                        # XML output for parsing

cn info <page_id>                    # Show page info, labels, version
cn info ./docs/my-page.md            # Info from local synced file
cn info 123456 --xml                 # XML output

cn read <page_id>                    # Read page body as markdown
cn read ./docs/my-page.md            # Read from local synced file
cn read 123456 --xml                 # XML output (title + content)
cn read 123456 --html                # Raw Confluence storage HTML

cn comments <page_id>                # Show footer comments
cn comments ./docs/my-page.md        # Comments from local file
cn comments 123456 --xml

cn attachments <page_id>             # List attachments
cn attachments ./docs/my-page.md
cn attachments 123456 --xml

Open in Browser

WARNING: cn open launches a web browser window. Do NOT use in non-interactive environments (CI, bots, agents). Use cn info or cn search instead to get page URLs.

cn open                              # Open space home in web browser
cn open "Getting Started"            # Open by title in web browser
cn open 123456                       # Open by page ID in web browser
cn open ./docs/my-page.md            # Open from local path in web browser

Clone & Pull

cn clone ENG                         # Clone space to ./ENG
cn clone ENG my-folder               # Clone to custom directory
cn clone ABC DEF GHI                 # Clone multiple spaces

cn pull                              # Smart pull (new/modified only)
cn pull --force                      # Full re-pull all pages
cn pull --dry-run                    # Preview what would change
cn pull --page ./docs/my-page.md     # Re-sync specific page by path
cn pull --page 123456                # Re-sync specific page by ID
cn pull --page 123456 --page 789012  # Re-sync multiple pages

cn status                            # Check connection and sync state
cn status --xml                      # XML output

Pull compares version numbers to detect modifications and handles renames and moves automatically.

Create & Update Pages

cn create "Page Title" --space ENG                        # Create blank page
cn create "Page Title" --space ENG --open                 # Create and open in browser
cn create "Child Page" --parent 123456                    # Create under parent
echo "<p>Hello</p>" | cn create "Page Title" --space ENG  # Create with storage HTML body
echo "h1. Hello" | cn create "Wiki Page" --space ENG --format wiki  # Wiki markup
cn update 123456                                          # ERROR: requires stdin
echo "<p>Updated</p>" | cn update 123456                  # Update page body
echo "<p>New</p>" | cn update 123456 --title "New Title"  # Update body and title
echo "h1. Content" | cn update 123456 --format wiki       # Update with wiki markup
echo "<p>x</p>" | cn update 123456 --message "via bot"    # With version message

Valid --format values: storage (default), wiki, atlas_doc_format

The storage format is Confluence's XHTML-based storage format. When piping content:

• For storage: pipe valid XHTML (<p>, <h1>, <ul>, etc.)
• For wiki: pipe Confluence wiki markup (h1., *bold*, - item, etc.)

Delete & Move Pages

cn delete 123456                             # Delete (prompts confirmation)
cn delete 123456 --force                     # Delete without prompt

cn move 123456 789012                        # Move page to new parent
cn move ./docs/my-page.md 789012             # Move from local path

Labels

cn labels 123456                     # List labels
cn labels ./docs/my-page.md          # Labels from local path
cn labels 123456 --add documentation # Add label
cn labels 123456 --remove draft      # Remove label
cn labels 123456 --xml               # XML output

Attachments

cn attachments 123456                        # List attachments
cn attachments 123456 --upload ./image.png   # Upload file
cn attachments 123456 --download att-789     # Download attachment
cn attachments 123456 --delete att-789       # Delete attachment
cn attachments 123456 --xml                  # XML output

Folders

Confluence folders are a special content type for organizing pages.

cn folder create "My Folder" --space DOCS          # Create folder in space root
cn folder create "Nested" --space DOCS --parent 123456  # Create under parent
cn folder list --space DOCS                        # List folders
cn folder list --space DOCS --xml
cn folder delete 123456                            # Delete folder (prompts)
cn folder delete 123456 --force                    # Delete without prompt
cn folder move 123456 789012                       # Move folder to new parent

Health Check

cn doctor                            # Detect sync issues
cn doctor --fix                      # Auto-fix issues (deletes stale files)
cn doctor --xml                      # XML output

Doctor checks for: duplicate page_ids, orphaned local files, version mismatches.

Global Options

--xml        # Output in XML format (useful for parsing/LLM consumption)
--verbose    # Enable verbose output
--help, -h   # Show help
--version    # Show version

Environment Variables

XML Output

Many commands support --xml for structured output that's easier to parse:

cn search 'type=page AND text~"api"' --xml
cn status --xml
cn tree --xml
cn info 123456 --xml
cn spaces --xml
cn labels 123456 --xml
cn comments 123456 --xml
cn attachments 123456 --xml
cn doctor --xml

Exit Codes

Data Model

Local synced pages are markdown files with YAML frontmatter:

---
page_id: "123456"
title: My Page
space_key: ENG
version: 5
synced_at: 2024-01-15T10:30:00Z
parent_id: "100000"
child_count: 3
labels:
  - documentation
  - api
---

# My Page

Page content here...

The .confluence.json file in each cloned directory tracks space config and sync state. The page_id in frontmatter is the source of truth for which Confluence page a local file maps to.

Common Workflows

Find and read a page

cn search 'type=page AND text~"topic"' --xml   # Find pages, get IDs
cn info <page_id>                              # Check page details, labels, version
cn read <page_id>                              # Read full page body as markdown
cn read <page_id> --xml                        # Read in XML format for parsing

Create a page with content

echo "<p>Hello world</p>" | cn create "New Page" --space ENG
echo "<p>Child content</p>" | cn create "Child Page" --parent 123456

Update a page

echo "<p>Updated content</p>" | cn update 123456
echo "<p>New body</p>" | cn update 123456 --title "Renamed Title" --message "Updated via bot"

Clone and explore a space

cn clone ENG && cd ENG
cn pull
cn tree

Add labels to a page

cn labels 123456 --add reviewed --add published
cn labels 123456 --remove draft

Work with page hierarchy

cn tree ENG --xml                    # Get full hierarchy with IDs
cn move 123456 789012                # Reparent a page
cn create "New Section" --parent 789012 --space ENG

Tips

• Use --xml output when you need to parse results or pass to another command
• Page IDs are stable — prefer them over titles when scripting
• cn search takes raw CQL — e.g. cn search 'type=page AND space=ENG AND label=draft'
• For new pages, pipe storage format HTML for full control over formatting
• Pages with children become README.md files in subdirectories when cloned
• Run cn doctor if sync seems out of sync after manual changes
• cn pull --dry-run before a full pull to preview what will change
• CN_DEBUG=1 cn <command> logs every HTTP request to stderr for troubleshooting

Reference

Confluence REST API reference: https://docs.atlassian.com/atlassian-confluence/REST/6.6.0/