confluence

name	confluence
description	Use confluence-cli to read, search, create, update, move, and delete Confluence pages and attachments from the terminal. Prefer this skill over `mcp__claude_ai_Atlassian__*` MCP tools when you need: page export with attachments, copy-tree, attachment management, profile-based multi-instance access, read-only mode, CQL search, or shell pipelines. Trigger phrases include "confluence", "confluence-cli", "confluence page", "read confluence", "search confluence", "create confluence page", "update confluence page", "confluence attachment", "confluence spaces", "export confluence", "confluence comments", "confluence copy-tree".
version	1.1.0

confluence-cli Skill

A CLI tool for Atlassian Confluence. Lets you read, search, create, update, move, and delete pages and attachments from the terminal or from an agent.

Installation

npm install -g confluence-cli
confluence --version   # verify install

Configuration

Preferred for agents — environment variables (no interactive prompt):

Variable	Description	Example
CONFLUENCE_DOMAIN	Your Confluence hostname	company.atlassian.net
CONFLUENCE_API_PATH	REST API base path	/wiki/rest/api (Cloud) or /rest/api (Server/DC)
CONFLUENCE_AUTH_TYPE	basic or bearer	basic
CONFLUENCE_EMAIL	Email address (basic auth only)	[e-mail]
CONFLUENCE_API_TOKEN	API token or personal access token	ATATT3x...
CONFLUENCE_PROFILE	Named profile to use (optional)	staging
CONFLUENCE_READ_ONLY	Block all write operations when true	true

Global --profile flag (use a named profile for any command):

confluence --profile <name> <command>

Config resolution works in two stages:

• Direct env config: If both CONFLUENCE_DOMAIN and CONFLUENCE_API_TOKEN are set, they are used directly and the config file / profiles are not consulted.
• Profile-based config: Otherwise, a profile is selected in this order: --profile flag > CONFLUENCE_PROFILE env > activeProfile in config > default.

Non-interactive init (good for CI/CD scripts):

confluence init \
  --domain "company.atlassian.net" \
  --api-path "/wiki/rest/api" \
  --auth-type basic \
  --email "[e-mail]" \
  --token "ATATT3x..."

Cloud vs Server/DC:

• Atlassian Cloud (*.atlassian.net): use --api-path "/wiki/rest/api", auth type basic with email + API token
• Atlassian Cloud (scoped token): use --domain "api.atlassian.com", --api-path "/ex/confluence/<your-cloud-id>/wiki/rest/api", auth type basic with email + scoped token. Get your Cloud ID from https://<your-site>.atlassian.net/_edge/tenant_info. Recommended for agents (least privilege).
• Self-hosted / Data Center: use --api-path "/rest/api", auth type bearer with a personal access token (no email needed)

Scoped API token for agents (recommended):

export CONFLUENCE_DOMAIN="api.atlassian.com"
export CONFLUENCE_API_PATH="/ex/confluence/<your-cloud-id>/wiki/rest/api"
export CONFLUENCE_AUTH_TYPE="basic"
export CONFLUENCE_EMAIL="[e-mail]"
export CONFLUENCE_API_TOKEN="your-scoped-token"

Required classic scopes for scoped tokens:

• Read-only: read:confluence-content.all, read:confluence-content.summary, read:confluence-space.summary, search:confluence
• Write: add write:confluence-content, write:confluence-file, write:confluence-space
• Attachments: readonly:content.attachment:confluence (download), write:confluence-file (upload)

Read-only mode (recommended for AI agents):

Prevents all write operations (create, update, delete, move, etc.) at the profile level. Useful when giving an AI agent access to Confluence for reading only.

# Via profile flag
confluence profile add agent --domain "company.atlassian.net" --token "xxx" --read-only

# Via environment variable (overrides config file)
export CONFLUENCE_READ_ONLY=true

When read-only mode is active, any write command exits with an error:

Error: This profile is in read-only mode. Write operations are not allowed.

profile list shows read-only profiles with a [read-only] badge.

---

Page ID Resolution

Most commands accept <pageId> — a numeric ID or any of the supported URL formats below.

Supported formats:

Format	Example
Numeric ID	123456789
?pageId= URL	https://company.atlassian.net/wiki/viewpage.action?pageId=123456789
Pretty /pages/<id> URL	https://company.atlassian.net/wiki/spaces/SPACE/pages/123456789/Page+Title
Display /display/<space>/<title> URL	https://company.atlassian.net/wiki/display/SPACE/Page+Title

confluence read 123456789
confluence read "https://company.atlassian.net/wiki/viewpage.action?pageId=123456789"
confluence read "https://company.atlassian.net/wiki/spaces/MYSPACE/pages/123456789/My+Page"

Note: Display-style URLs (/display/<space>/<title>) perform a title-based lookup, so the page title in the URL must match exactly. When possible, prefer numeric IDs or /pages/<id> URLs for reliability.

Content Formats

Format	Notes
markdown	Recommended for agent-generated content. Automatically converted by the API.
storage	Confluence XML storage format (default for create/update). Use for programmatic round-trips.
html	Raw HTML.
text	Plain text — for read/export output only, not for creation.

---

Constraints

• Always use confluence CLI — it must be globally installed via npm.
• Configuration must already be set up (env vars or confluence init). If not, guide the user through setup.
• For simple Confluence read/create/update operations where mcp__claude_ai_Atlassian__* MCP tools suffice, prefer MCP (less overhead, no shell spawning). Switch to CLI for features MCP lacks.

Preflight Check

Before running any confluence command, verify the installation:

confluence --version

If this fails:

• "command not found" — confluence-cli is not installed. Suggest: npm install -g confluence-cli
• "No configuration found" — Config not set up. Suggest: confluence init or set env vars (CONFLUENCE_DOMAIN, CONFLUENCE_API_TOKEN)

When to Use CLI vs. MCP

Capability	mcp__claude_ai_Atlassian__* MCP	confluence CLI
Read page	Yes — getConfluencePage	Yes — confluence read
Search (CQL)	Yes — searchConfluenceUsingCql	Yes — confluence search
Create page	Yes — createConfluencePage	Yes — confluence create
Update page	Yes — updateConfluencePage	Yes — confluence update
Get comments	Yes — various comment tools	Yes — confluence comments
Add comment	Yes — createConfluenceFooterComment	Yes — confluence comment
List spaces	Yes — getConfluenceSpaces	Yes — confluence spaces
Page descendants	Yes — getConfluencePageDescendants	Yes — confluence children
Export page + attachments	No	Yes — confluence export
Copy page tree	No	Yes — confluence copy-tree
Move page	No	Yes — confluence move
Delete page	No	Yes — confluence delete
Attachment upload/download	No	Yes — confluence attachment-upload / attachments --download
Attachment delete	No	Yes — confluence attachment-delete
Edit raw storage XML	No	Yes — confluence edit

Rule of thumb: Use MCP for simple read/create/update. Use CLI for export, copy-tree, attachments, move, delete, multi-instance, and read-only mode.

Quick Reference

Action	Command
Read page	confluence read <pageId> --format markdown
Page metadata	confluence info <pageId>
Find by title	confluence find "Title" --space SPACE
Search	confluence search "query" --limit 20
List spaces	confluence spaces
List children	confluence children <pageId> --recursive --format tree
Create page	confluence create "Title" SPACE --file ./content.md --format markdown
Create child	confluence create-child "Title" <parentId> --content "# Hello" --format markdown
Update page	confluence update <pageId> --file ./updated.md --format markdown
Move page	confluence move <pageId> <newParentId>
Delete page	confluence delete <pageId> --yes
Export page	confluence export <pageId> --format markdown --dest ./docs
Copy tree	confluence copy-tree <sourceId> <targetParentId> --dry-run
List attachments	confluence attachments <pageId>
Upload attachment	confluence attachment-upload <pageId> --file ./report.pdf
Download attachments	confluence attachments <pageId> --download --dest ./downloads
List comments	confluence comments <pageId> --format json --all
Add comment	confluence comment <pageId> --content "Note" --location footer
Switch profile	confluence profile use staging

---

Commands Reference

init

Initialize configuration. Saves credentials to ~/.confluence-cli/config.json.

confluence init [--domain <domain>] [--api-path <path>] [--auth-type basic|bearer] [--email <email>] [--token <token>] [--read-only]

All flags are optional; omitting any flag triggers an interactive prompt for that field. Provide all flags to run fully non-interactive. Use the global --profile flag to save to a named profile:

confluence --profile staging init --domain "staging.example.com" --auth-type bearer --token "your-token"

---

read <pageId>

Read page content. Outputs to stdout.

confluence read <pageId> [--format html|text|markdown]

Option	Default	Description
--format	text	Output format: html, text, or markdown

confluence read 123456789
confluence read 123456789 --format markdown

---

info <pageId>

Get page metadata (title, ID, type, status, space).

confluence info <pageId>

confluence info 123456789

---

find <title>

Find a page by exact or partial title. Returns the first match.

confluence find <title> [--space <spaceKey>]

Option	Description
--space	Restrict search to a specific space key

confluence find "Architecture Overview"
confluence find "API Reference" --space MYSPACE

---

search <query>

Search pages using a keyword or CQL expression.

confluence search <query> [--limit <number>]

Option	Default	Description
--limit	10	Maximum number of results

confluence search "deployment pipeline"
confluence search "type=page AND space=MYSPACE" --limit 50

---

spaces

List all accessible Confluence spaces (key and name).

confluence spaces

---

children <pageId>

List child pages of a page.

confluence children <pageId> [--recursive] [--max-depth <number>] [--format list|tree|json] [--show-id] [--show-url]

Option	Default	Description
--recursive	false	List all descendants recursively
--max-depth	10	Maximum depth for recursive listing
--format	list	Output format: list, tree, or json
--show-id	false	Show page IDs
--show-url	false	Show page URLs

confluence children 123456789
confluence children 123456789 --recursive --format json
confluence children 123456789 --recursive --format tree --show-id

---

create <title> <spaceKey>

Create a new top-level page in a space.

confluence create <title> <spaceKey> [--content <string>] [--file <path>] [--format storage|html|markdown]

Option	Default	Description
--content	—	Inline content string
--file	—	Path to content file
--format	storage	Content format

Either --content or --file is required.

confluence create "Project Overview" MYSPACE --content "# Hello" --format markdown
confluence create "Release Notes" MYSPACE --file ./notes.md --format markdown

Outputs the new page ID and URL on success.

---

create-child <title> <parentId>

Create a child page under an existing page. Inherits the parent's space automatically.

confluence create-child <title> <parentId> [--content <string>] [--file <path>] [--format storage|html|markdown]

Options are identical to create. Either --content or --file is required.

confluence create-child "Chapter 1" 123456789 --content "Content here" --format markdown
confluence create-child "API Guide" 123456789 --file ./api.md --format markdown

---

update <pageId>

Update an existing page's title and/or content. At least one of --title, --content, or --file is required.

confluence update <pageId> [--title <title>] [--content <string>] [--file <path>] [--format storage|html|markdown]

Option	Default	Description
--title	—	New title
--content	—	Inline content string
--file	—	Path to content file
--format	storage	Content format

confluence update 123456789 --title "New Title"
confluence update 123456789 --file ./updated.md --format markdown
confluence update 123456789 --title "New Title" --file ./updated.xml --format storage

---

move <pageId_or_url> <newParentId_or_url>

Move a page to a new parent. Both pages must be in the same space.

confluence move <pageId_or_url> <newParentId_or_url> [--title <newTitle>]

Option	Description
--title	Rename the page during the move

confluence move 123456789 987654321
confluence move 123456789 987654321 --title "Renamed After Move"

---

delete <pageIdOrUrl>

Delete (trash) a page by ID or URL.

confluence delete <pageIdOrUrl> [--yes]

Option	Description
--yes	Skip confirmation prompt (required for non-interactive/agent use)

confluence delete 123456789 --yes

---

edit <pageId>

Fetch a page's raw storage-format content for editing locally.

confluence edit <pageId> [--output <file>]

Option	Description
--output	Save content to a file (instead of printing to stdout)

confluence edit 123456789 --output ./page.xml
# Edit page.xml, then:
confluence update 123456789 --file ./page.xml --format storage

---

export <pageId>

Export a page and its attachments to a local directory.

confluence export <pageId> [--format html|text|markdown] [--dest <directory>] [--file <filename>] [--attachments-dir <name>] [--pattern <glob>] [--referenced-only] [--skip-attachments]

Option	Default	Description
--format	markdown	Content format for the exported file
--dest	.	Base directory to export into
--file	page.<ext>	Filename for the content file
--attachments-dir	attachments	Subdirectory name for attachments
--pattern	—	Glob filter for attachments (e.g. *.png)
--referenced-only	false	Only download attachments referenced in the page content
--skip-attachments	false	Do not download attachments

confluence export 123456789 --format markdown --dest ./docs
confluence export 123456789 --format markdown --dest ./docs --skip-attachments
confluence export 123456789 --pattern "*.png" --dest ./output

Creates a subdirectory named after the page title under --dest.

---

attachments <pageId>

List or download attachments for a page.

confluence attachments <pageId> [--limit <n>] [--pattern <glob>] [--download] [--dest <directory>]

Option	Default	Description
--limit	all	Maximum number of attachments to fetch
--pattern	—	Filter by filename glob (e.g. *.pdf)
--download	false	Download matching attachments
--dest	.	Directory to save downloads

confluence attachments 123456789
confluence attachments 123456789 --pattern "*.pdf" --download --dest ./downloads

---

attachment-upload <pageId>

Upload one or more files to a page. --file can be repeated for multiple files.

confluence attachment-upload <pageId> --file <path> [--file <path> ...] [--comment <text>] [--replace] [--minor-edit]

Option	Description
--file	File to upload (required, repeatable)
--comment	Comment for the attachment(s)
--replace	Replace an existing attachment with the same filename
--minor-edit	Mark the upload as a minor edit

confluence attachment-upload 123456789 --file ./report.pdf
confluence attachment-upload 123456789 --file ./a.png --file ./b.png --replace

---

attachment-delete <pageId> <attachmentId>

Delete an attachment from a page.

confluence attachment-delete <pageId> <attachmentId> [--yes]

Option	Description
--yes	Skip confirmation prompt

confluence attachment-delete 123456789 att-987 --yes

---

comments <pageId>

List comments for a page.

confluence comments <pageId> [--format text|markdown|json] [--limit <n>] [--start <n>] [--location inline,footer,resolved] [--depth all] [--all]

Option	Default	Description
--format	text	Output format: text, markdown, or json
--limit	25	Maximum comments per page
--start	0	Start index for pagination
--location	—	Filter by location: inline, footer, resolved (comma-separated)
--depth	—	Leave empty for root-only; all for all nested replies
--all	false	Fetch all comments (ignores pagination)

confluence comments 123456789
confluence comments 123456789 --format json --all
confluence comments 123456789 --location footer --depth all

---

comment <pageId>

Create a comment on a page (footer or inline).

confluence comment <pageId> [--content <string>] [--file <path>] [--format storage|html|markdown] [--parent <commentId>] [--location footer|inline] [--inline-selection <text>] [--inline-original-selection <text>] [--inline-marker-ref <ref>] [--inline-properties <json>]

Option	Default	Description
--content	—	Inline content string
--file	—	Path to content file
--format	storage	Content format
--parent	—	Reply to a comment by ID
--location	footer	footer or inline
--inline-selection	—	Highlighted selection text (inline only)
--inline-original-selection	—	Original selection text (inline only)
--inline-marker-ref	—	Marker reference (inline only)
--inline-properties	—	Full inline properties as JSON (advanced)

Either --content or --file is required.

confluence comment 123456789 --content "Looks good!" --location footer
confluence comment 123456789 --content "See note" --parent 456 --location footer

Note on inline comments: Creating a brand-new inline comment requires editor highlight metadata (matchIndex, lastFetchTime, serializedHighlights) that is only available in the Confluence editor. This metadata is not accessible via the REST API, so inline comment creation will typically fail with a 400 error. Use --location footer or reply to an existing inline comment with --parent <commentId> instead.

---

comment-delete <commentId>

Delete a comment by its ID.

confluence comment-delete <commentId> [--yes]

Option	Description
--yes	Skip confirmation prompt

confluence comment-delete 456789 --yes

---

copy-tree <sourcePageId> <targetParentId> [newTitle]

Copy a page and all its children to a new location.

confluence copy-tree <sourcePageId> <targetParentId> [newTitle] [--max-depth <depth>] [--exclude <patterns>] [--delay-ms <ms>] [--copy-suffix <suffix>] [--dry-run] [--fail-on-error] [--quiet]

Option	Default	Description
--max-depth	10	Maximum depth to copy
--exclude	—	Comma-separated title patterns to exclude (supports wildcards)
--delay-ms	100	Delay between sibling creations in ms
--copy-suffix	" (Copy)"	Suffix appended to the root page title
--dry-run	false	Preview operations without creating pages
--fail-on-error	false	Exit with non-zero code if any page fails
--quiet	false	Suppress progress output

# Preview first
confluence copy-tree 123456789 987654321 --dry-run

# Copy with a custom title
confluence copy-tree 123456789 987654321 "Backup Copy"

# Copy excluding certain pages
confluence copy-tree 123456789 987654321 --exclude "Draft*,Archive*"

---

profile list

List all configuration profiles with the active profile marked.

confluence profile list

---

profile use <name>

Switch the active configuration profile.

confluence profile use <name>

confluence profile use staging

---

profile add <name>

Add a new configuration profile. Supports the same options as init (interactive, non-interactive, or hybrid).

confluence profile add <name> [--domain <domain>] [--api-path <path>] [--auth-type basic|bearer] [--email <email>] [--token <token>] [--protocol http|https] [--read-only]

Profile names may contain letters, numbers, hyphens, and underscores only.

confluence profile add staging --domain "staging.example.com" --auth-type bearer --token "xyz"

---

profile remove <name>

Remove a configuration profile (prompts for confirmation). Cannot remove the only remaining profile.

confluence profile remove <name>

confluence profile remove staging

---

stats

Show local usage statistics.

confluence stats

---

install-skill

Copy the Claude Code skill documentation into your project's .claude/skills/ directory so Claude Code can learn confluence-cli automatically.

confluence install-skill [--dest <directory>] [--yes]

Option	Default	Description
--dest	./.claude/skills/confluence	Target directory
--yes	false	Skip overwrite confirmation

confluence install-skill

---

Common Agent Workflows

Read → Edit → Update (round-trip)

# 1. Fetch raw storage XML
confluence edit 123456789 --output ./page.xml

# 2. Modify page.xml with your tool of choice

# 3. Push the updated content
confluence update 123456789 --file ./page.xml --format storage

Surgical edits on storage XML — safety rules

When scripting targeted replacements on a fetched page.xml:

1. Re-fetch immediately before every update. Pages get edited concurrently (humans, other agent sessions). confluence edit prints the current version — if any time passed since your fetch, fetch again and re-apply the edits on the fresh copy; otherwise update silently clobbers the newer version.

2. Never anchor replacements on local-id attributes. The Confluence editor normalizes or strips local-ids whenever a human edits the page, so id-based anchors break without warning. Anchor on unique literal text instead.

3. Assert uniqueness per replacement so a drifted anchor fails loudly instead of corrupting the page:

   def rep(old, new):
       global src
       assert src.count(old) == 1, f"anchor drift: {old[:60]!r}"
       src = src.replace(old, new, 1)
   

4. New elements you insert may omit local-id — Confluence assigns ids on save.

Build a documentation hierarchy

# Create root page, note the returned ID (e.g. 111222333)
confluence create "Project Overview" MYSPACE --content "# Overview" --format markdown

# Add children under it
confluence create-child "Architecture" 111222333 --content "# Architecture" --format markdown
confluence create-child "API Reference" 111222333 --file ./api.md --format markdown
confluence create-child "Runbooks" 111222333 --content "# Runbooks" --format markdown

Copy a full page tree

# Preview first
confluence copy-tree 123456789 987654321 --dry-run

# Execute the copy
confluence copy-tree 123456789 987654321 "Backup Copy"

Export a page for local editing

confluence export 123456789 --format markdown --dest ./local-docs
# => ./local-docs/<page-title>/page.md + ./local-docs/<page-title>/attachments/

Process children as JSON

confluence children 123456789 --recursive --format json | jq '.[].id'

Search and process results

confluence search "release notes" --limit 20

---

Agent Tips

• Always use --yes on destructive commands (delete, comment-delete, attachment-delete) to avoid interactive prompts blocking the agent.
• Prefer --format markdown when creating or updating content from agent-generated text — it's the most natural format and the API converts it automatically.
• Use --format json on children and comments for machine-parseable output.
• ANSI color codes: stdout may contain ANSI escape sequences. Pipe through | cat or use NO_COLOR=1 if your downstream tool doesn't handle them.
• Page ID vs URL: when you have a Confluence URL, extract ?pageId=<number> and pass the number. Do not pass pretty/display URLs — they are not supported.
• Cross-space moves: confluence move only works within the same space. Moving across spaces is not supported.
• Multiple instances: Use --profile <name> or CONFLUENCE_PROFILE env var to target different Confluence instances without reconfiguring.
• Read-only mode: Set CONFLUENCE_READ_ONLY=true or use --read-only when creating profiles to prevent accidental writes. This is enforced at the CLI level — all write commands will be blocked.

Error Patterns

Error	Cause	Fix
No configuration found	No config file and no env vars set	Set env vars or run confluence init
Cross-space moves are not supported	move used across spaces	Copy with copy-tree instead
400 on inline comment creation	Editor metadata required	Use --location footer or reply to existing inline comment with --parent
File not found: <path>	--file path doesn't exist	Check the path before calling the command
At least one of --title, --file, or --content must be provided	update called with no content options	Provide at least one of the required options
Profile "<name>" not found!	Specified profile doesn't exist	Run confluence profile list to see available profiles
Cannot delete the only remaining profile.	Tried to remove the last profile	Add another profile before removing
This profile is in read-only mode	Write command used with a read-only profile	Use a writable profile or remove readOnly from config

Behavior Scenarios

Scenario 1: Read a Confluence page

Given confluence-cli is installed and configured
When the user asks to read a Confluence page (by ID or URL)
Then run `confluence read <pageId> --format markdown`
And present the page content

Scenario 2: Search Confluence

Given confluence-cli is installed and configured
When the user asks to search Confluence for a term
Then run `confluence search "<term>" --limit 10`
And present matching pages with titles and IDs

Scenario 3: Create or update a page

Given confluence-cli is installed and configured
When the user asks to create or update a Confluence page
Then prefer `--format markdown` for agent-generated content
And use `--file` for larger content (write to a temp file first)
And report the page ID and URL on success

Scenario 4: Export a page tree

Given confluence-cli is installed and configured
When the user asks to export a Confluence page
Then run `confluence export <pageId> --format markdown --dest <directory>`
And report the exported files and attachment count

Scenario 5: confluence-cli not installed

Given confluence-cli is not installed or not configured
When the user asks to perform any Confluence CLI operation
Then check with `confluence --version`
And guide the user to install: `npm install -g confluence-cli`
And guide config setup: `confluence init` or set env vars

Scenario 6: Multi-instance operation

Given the user works with multiple Confluence instances
When the user specifies an instance or profile
Then use the `--profile <name>` flag on the command
And if no profile exists, guide setup with `confluence profile add <name>`