Allows managing Confluence Cloud wiki pages, spaces, blog posts, comments, attachments, labels, tasks, and content properties via the Confluence REST API v2.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@confluence-mcp list pages in space DEV"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP server for administering Atlassian Confluence Cloud wiki pages. Built with FastMCP and the Confluence REST API v2.
Create, read, update, delete, and restore pages
Move (reparent) pages within a space; transfer page ownership
Navigate the page tree (children, ancestors, descendants)
Bulk-fetch multiple pages in a single request
Create, read, update, and delete spaces; manage space content properties
Full blog post CRUD (create, read, update, delete)
Read labels (list all labels, list pages/blogposts by label)
Inspect version history
Manage footer and inline comments (including creating inline comments)
List and delete attachments
Manage page and space content properties (key/value metadata)
Manage tasks
Bulk user lookup (resolve multiple account IDs to profiles)
Related MCP server: inhouse_confluence_mcp
Python 3.11+
uv package manager
An Atlassian Confluence Cloud account with API access
git clone <repo-url>
cd confluence
uv syncCopy .env.example to .env and fill in your values:
cp .env.example .envVariable | Required | Description |
| Yes | Your Confluence base URL, e.g. |
| Basic Auth | Atlassian account email |
| Basic Auth | |
| Bearer Auth | Personal Access Token (alternative to Basic Auth) |
Use either Basic Auth (EMAIL + API_TOKEN) or a Bearer token — not both.
uv run confluence-mcpThe server communicates over stdio (standard MCP transport).
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or the equivalent on your OS:
{
"mcpServers": {
"confluence": {
"command": "uv",
"args": [
"--directory", "/path/to/confluence",
"run", "confluence-mcp"
],
"env": {
"CONFLUENCE_BASE_URL": "https://your-domain.atlassian.net",
"CONFLUENCE_EMAIL": "you@example.com",
"CONFLUENCE_API_TOKEN": "your-api-token"
}
}
}
}Add to your OpenCode MCP config:
{
"confluence": {
"command": "uv",
"args": [
"--directory", "/path/to/confluence",
"run", "confluence-mcp"
]
}
}Set the env vars in your shell or .env file.
Tool | Description |
| List pages filtered by space, title, status, or parent |
| Get a page by ID (with optional body format and version) |
| Fetch multiple pages by ID in a single request |
| Create a new page in a space |
| Update content, title, parent (move), or owner of a page |
| Rename a page without changing its body or version metadata |
| Move a page to trash (or permanently delete with |
| Restore a trashed page to |
| List permitted operations for a page |
Moving a page: pass parent_id in update_page to reparent within the same space.
Cross-space moves are not supported by the Confluence v2 API.
Transferring ownership: pass owner_id (Atlassian account ID) in update_page.
Tool | Description |
| All child pages (recursive, paginated) |
| Immediate children only |
| Full breadcrumb trail to the root |
| All descendants |
Tool | Description |
| List spaces (filterable by type, status, key, label) |
| Get a space by ID |
| Create a new space |
| List permitted operations for a space |
| List all labels used on content within a space |
| List pages in a space |
| List blog posts in a space |
| List permission assignments for a space |
Tool | Description |
| List key/value properties on a space |
| Get a specific space property |
| Create a new space property |
| Update an existing space property |
| Delete a space property |
Tool | Description |
| List blog posts (filterable by space, status) |
| Get a blog post by ID |
| Create a new blog post in a space |
| Update content, title, or status of a blog post |
| Move a blog post to trash (or permanently delete with |
Label add/remove is available via the Confluence v1 API only; the v2 API is read-only for labels.
Tool | Description |
| List all labels across the Confluence instance |
| Labels attached to a page |
| Labels used across a space |
| Pages that carry a specific label |
| Blog posts that carry a specific label |
Tool | Description |
| Version history of a page |
| Details of a specific page version |
Tool | Description |
| Footer comments on a page |
| Get a footer comment by ID |
| Create a comment on a page, blog post, or as a reply |
| Update comment content |
| Delete a comment |
| List replies to a comment |
Tool | Description |
| Inline comments on a page |
| Get an inline comment by ID |
| Create an inline comment anchored to a text selection |
| Update content or resolved state |
| Delete an inline comment |
Tool | Description |
| List attachments on a page |
| Get an attachment by ID |
| Delete an attachment |
Uploading attachments requires multipart form data and is not supported by this server.
Tool | Description |
| List key/value properties on a page |
| Get a specific property |
| Create a new property |
| Update an existing property |
| Delete a property |
Tool | Description |
| List tasks (filterable by status, space, page, assignee) |
| Get a task by ID |
| Update task status, assignee, or due date |
Tool | Description |
| Resolve content IDs to their types (page, blogpost, etc.) |
| Resolve multiple Atlassian account IDs to user profiles |
Body formats: storage (XHTML, default for write), wiki (Confluence wiki markup), atlas_doc_format (Atlassian Document Format / ADF), view (rendered HTML, read-only).
Versioning: update_page, update_blogpost, update_footer_comment, and update_inline_comment all require passing the current version_number. The API increments it automatically. Always call get_page / get_blogpost first to obtain the current version.
Labels: To add or remove labels, use the Confluence v1 REST API (/wiki/rest/api/content/{id}/label) directly — this is not exposed in this server.
Fetch multiple pages by their IDs in a single request.
Returns a results list with one entry per resolved page ID.
Args:
page_ids: List of page IDs to retrieve (up to 250).
body_format: Body representation — 'storage', 'atlas_doc_format', or 'view'.
Inputs: page_ids (required), body_format.
Resolve multiple Atlassian account IDs to user profiles in a single call.
Useful when you have a list of assignee/author IDs from tasks or comments
and want to retrieve display names and emails in bulk.
Args:
account_ids: List of Atlassian account IDs to look up (up to 200).
Inputs: account_ids (required).
Convert a list of content IDs to their content types (page, blogpost, etc.).
Useful when you have an ID but don't know whether it refers to a page,
blog post, attachment, or other content type.
Args:
content_ids: List of content IDs to look up.
Inputs: content_ids (required).
Create a new blog post in a space.
Args:
space_id: The ID of the space to publish in.
title: Title of the blog post.
body_value: HTML/storage body content.
body_representation: Body format — 'storage', 'wiki', or 'atlas_doc_format'.
status: 'current' to publish immediately, or 'draft' to save as draft.
Inputs: title (required), space_id (required), body_value (required), status, +1 more.
Create an inline comment anchored to a text selection on a page.
Args:
page_id: The ID of the page to comment on.
body_value: HTML/storage-format body of the comment.
inline_marker_ref: The marker reference identifying the text selection
(obtained from the page's inline marker data).
body_representation: Body format — 'storage', 'wiki', or 'atlas_doc_format'.
resolved: Whether to create the comment in a resolved state.
Inputs: page_id (required), body_value (required), inline_marker_ref (required), resolved, +1 more.
Create a new page in a space.
Args:
space_id: Target space ID.
title: Page title.
body_value: Page body content.
body_representation: Format of body_value — storage (default, XHTML), wiki, atlas_doc_format.
parent_id: Parent page ID (omit to create at space root).
status: 'current' (published, default) or 'draft'.
Inputs: title (required), space_id (required), body_value (required), status, +2 more.
Create a content property on a page.
Args:
page_id: Page ID.
key: Property key (must be unique on the page).
value: Property value — any JSON-serializable value.
Inputs: key (required), value (required), page_id (required).
Create a new Confluence space.
Args:
key: Unique space key (e.g. 'ENG'). Must be uppercase letters/numbers.
name: Human-readable space name.
description: Optional plain-text description of the space.
Inputs: key (required), name (required), description.
Create a new content property on a space.
Args:
space_id: The ID of the space.
key: Property key (must be unique within the space).
value: Property value (string, number, dict, or list).
Inputs: key (required), value (required), space_id (required).
Delete an attachment.
Args:
attachment_id: Attachment ID.
Inputs: attachment_id (required).
Delete a blog post (moves to trash, or permanently purges).
Args:
blogpost_id: The ID of the blog post to delete.
purge: If True, permanently delete instead of moving to trash.
Inputs: blogpost_id (required), purge.
Delete an inline comment.
Args:
comment_id: Comment ID.
Inputs: comment_id (required).
Delete a page (moves to trash by default).
To permanently delete, first call delete_page to trash the page, then call
delete_page again with purge=True.
Args:
page_id: ID of the page to delete.
purge: If True, permanently delete a trashed page. The page must already
be in the trash; calling purge=True on a current page returns 400.
Inputs: page_id (required), purge.
Delete a content property from a page.
Args:
page_id: Page ID.
property_id: Property ID.
Inputs: page_id (required), property_id (required).
Delete a content property from a space.
Args:
space_id: The ID of the space.
property_id: The ID of the property to delete.
Inputs: space_id (required), property_id (required).
Get an attachment by ID.
Args:
attachment_id: Attachment ID.
version: Specific version number.
Inputs: attachment_id (required), version.
Get a blog post by ID.
Args:
blogpost_id: The ID of the blog post.
body_format: Body representation — 'storage', 'atlas_doc_format', or 'view'.
version: Retrieve a specific historical version number.
Inputs: blogpost_id (required), version, body_format.
Get an inline comment by ID.
Args:
comment_id: Comment ID.
body_format: Body representation.
version: Specific version number.
Inputs: comment_id (required), version, body_format.
Get a page by ID.
Args:
page_id: Page ID.
body_format: Body representation — storage (XHTML), view, atlas_doc_format.
get_draft: If True, return the current draft instead of published version.
version: Return a specific historical version number.
Inputs: page_id (required), version, get_draft, body_format.
Get all ancestor pages (breadcrumb trail) for a page.
Args:
page_id: Page ID.
Inputs: page_id (required).
Get all child pages of a page (recursive, paginated).
Args:
page_id: Parent page ID.
cursor: Pagination cursor.
limit: Max results.
sort: Sort field.
Inputs: page_id (required), sort, limit, cursor.
Get all descendants of a page.
Args:
page_id: Page ID.
limit: Max results.
cursor: Pagination cursor.
Inputs: page_id (required), limit, cursor.
Get the immediate (one-level-deep) children of a page.
Args:
page_id: Parent page ID.
cursor: Pagination cursor.
limit: Max results.
Inputs: page_id (required), limit, cursor.
Get the permitted operations (read, update, delete, …) for a page.
Args:
page_id: Page ID.
Inputs: page_id (required).
Get a specific content property on a page.
Args:
page_id: Page ID.
property_id: Property ID.
Inputs: page_id (required), property_id (required).
Get details for a specific version of a page.
Args:
page_id: Page ID.
version_number: Version number to retrieve.
Inputs: page_id (required), version_number (required).
Get a space by its ID.
Args:
space_id: Space ID.
Inputs: space_id (required).
List all labels used on content within a space.
Args:
space_id: The ID of the space.
prefix: Filter by label prefix (e.g. 'global', 'my', 'team').
cursor: Pagination cursor from a previous response.
limit: Max results to return (default 25, max 250).
Inputs: space_id (required), limit, cursor, prefix.
Get the operations (permissions) available to the current user in a space.
Args:
space_id: The ID of the space.
Inputs: space_id (required).
Get permission assignments for a space.
Args:
space_id: Space ID.
cursor: Pagination cursor.
limit: Max results.
Inputs: space_id (required), limit, cursor.
Get a single content property of a space by its ID.
Args:
space_id: The ID of the space.
property_id: The ID of the property.
Inputs: space_id (required), property_id (required).
Get a task by ID.
Args:
task_id: Task ID.
body_format: Body representation for the task content.
Inputs: task_id (required), body_format.
List blog posts, optionally filtered by space.
Args:
space_id: Restrict to this space ID.
status: Filter by status — 'current' or 'draft'.
sort: Sort order (e.g. 'created-date', '-created-date', 'title').
cursor: Pagination cursor from a previous response.
limit: Max results to return (default 25, max 250).
Inputs: sort, limit, cursor, status, +1 more.
List blog posts in a space.
Args:
space_id: Space ID.
sort: Sort field.
status: Blog post status filter.
cursor: Pagination cursor.
limit: Max results.
Inputs: space_id (required), sort, limit, cursor, +1 more.
List blog posts that carry a specific label.
Args:
label_id: The ID of the label.
space_id: Restrict results to this space ID.
body_format: Body representation — 'storage', 'atlas_doc_format', or 'view'.
sort: Sort order (e.g. 'created-date', '-created-date', 'title').
cursor: Pagination cursor from a previous response.
limit: Max results to return (default 25, max 250).
Inputs: label_id (required), sort, limit, cursor, +2 more.
List labels across the entire Confluence instance.
Args:
label_prefix: Filter by prefix — 'global', 'my', or 'team'.
body_format: Body representation for associated content.
cursor: Pagination cursor from a previous response.
limit: Max results to return (default 25, max 250).
Inputs: limit, cursor, body_format, label_prefix.
List attachments on a page.
Args:
page_id: Page ID.
sort: Sort field.
cursor: Pagination cursor.
status: Attachment status filter.
mediatype: MIME type filter, e.g. 'image/png'.
filename: Filename filter.
limit: Max results.
Inputs: page_id (required), sort, limit, cursor, +3 more.
List inline comments on a page.
Args:
page_id: Page ID.
body_format: Body representation.
sort: Sort field.
cursor: Pagination cursor.
limit: Max results.
Inputs: page_id (required), sort, limit, cursor, +1 more.
List labels attached to a page.
Args:
page_id: Page ID.
prefix: Filter by label prefix (e.g. 'global', 'team').
sort: Sort field.
cursor: Pagination cursor.
limit: Max results.
Inputs: page_id (required), sort, limit, cursor, +1 more.
List content properties on a page.
Args:
page_id: Page ID.
key: Filter by property key.
cursor: Pagination cursor.
limit: Max results.
Inputs: page_id (required), key, limit, cursor.
List pages, optionally filtered by space, title, status, or parent.
Args:
space_id: Filter by space ID.
title: Exact title match.
status: Page status — current, archived, trashed, deleted.
ancestor_id: Filter by parent page ID.
sort: Sort field, e.g. 'title', '-title', 'created-date', '-modified-date'.
limit: Max results (1–250).
cursor: Pagination cursor from a previous response.
Inputs: sort, limit, title, cursor, +3 more.
List all pages in a space.
Args:
space_id: Space ID.
depth: 'all' (default) or 'root' (top-level pages only).
sort: Sort field.
status: Page status filter.
cursor: Pagination cursor.
limit: Max results.
Inputs: space_id (required), sort, depth, limit, +2 more.
List all pages that carry a specific label.
Args:
label_id: Label ID.
space_id: Narrow results to a specific space.
body_format: Body representation in the returned pages.
sort: Sort field.
cursor: Pagination cursor.
limit: Max results.
Inputs: label_id (required), sort, limit, cursor, +2 more.
List the version history of a page.
Args:
page_id: Page ID.
cursor: Pagination cursor.
limit: Max results.
sort: Sort field — 'created-date' (oldest first) or '-created-date' (newest first).
Inputs: page_id (required), sort, limit, cursor.
List labels used across a space.
Args:
space_id: Space ID.
prefix: Filter by label prefix.
sort: Sort field.
cursor: Pagination cursor.
limit: Max results.
Inputs: space_id (required), sort, limit, cursor, +1 more.
List content properties of a space.
Args:
space_id: The ID of the space.
key: Filter by property key.
cursor: Pagination cursor from a previous response.
limit: Max results to return.
Inputs: space_id (required), key, limit, cursor.
List Confluence spaces.
Args:
ids: Comma-separated space IDs to filter by.
keys: Comma-separated space keys to filter by.
type: Space type — 'global' or 'personal'.
status: Space status — 'current' or 'archived'.
labels: Comma-separated labels to filter by.
sort: Sort field, e.g. 'name', '-name', 'key'.
cursor: Pagination cursor.
limit: Max results.
Inputs: ids, keys, sort, type, +4 more.
List tasks in Confluence.
Args:
status: Task status — 'open' or 'complete'.
space_id: Filter by space ID.
page_id: Filter by page ID.
assigned_to: Filter by assignee account ID.
created_by: Filter by creator account ID.
due_at_from: ISO 8601 — tasks due on or after this date.
due_at_to: ISO 8601 — tasks due on or before this date.
cursor: Pagination cursor.
limit: Max results.
Inputs: limit, cursor, status, page_id, +5 more.
Restore a trashed page by setting its status back to 'current'.
Args:
page_id: ID of the trashed page.
title: Page title (unchanged or new).
body_value: Page body content (unchanged or new).
version_number: The next version number to store — must be current version + 1.
body_representation: Format of body_value — storage (default), wiki, atlas_doc_format.
parent_id: Parent page ID (required if the original parent was also deleted).
Inputs: title (required), page_id (required), body_value (required), version_number (required), +2 more.
Update an existing blog post.
Args:
blogpost_id: The ID of the blog post to update.
title: New title.
body_value: New body content.
version_number: The *new* version number (current version + 1).
body_representation: Body format — 'storage', 'wiki', or 'atlas_doc_format'.
status: 'current' or 'draft'.
Inputs: title (required), body_value (required), blogpost_id (required), version_number (required), +2 more.
Update an inline comment (content or resolved state).
Args:
comment_id: Comment ID.
body_value: New comment content.
version_number: The next version number to store — must be current version + 1.
body_representation: Format of body_value — storage (default), wiki, atlas_doc_format.
resolved: Set to True to mark the comment thread as resolved, False to reopen.
Inputs: body_value (required), comment_id (required), version_number (required), resolved, +1 more.
Update an existing page (content, title, parent, or owner).
To move a page to a different parent within the same space, pass the new
parent_id. Cross-space moves are not supported by the v2 API.
To transfer ownership, pass the new owner's account ID as owner_id.
Args:
page_id: ID of the page to update.
title: New (or unchanged) title.
body_value: New (or unchanged) body content.
version_number: The next version number to store — must be current version + 1.
body_representation: Format of body_value — storage (default), wiki, atlas_doc_format.
status: 'current' (published) or 'draft'.
parent_id: New parent page ID to move/reparent the page (same space only).
owner_id: Atlassian account ID of the new page owner.
version_message: Optional commit message stored in the version history.
Inputs: title (required), page_id (required), body_value (required), version_number (required), +5 more.
Update a content property on a page.
Args:
page_id: Page ID.
property_id: Property ID.
key: Property key.
value: New property value.
version_number: The next version number to store — must be current version + 1.
Inputs: key (required), value (required), page_id (required), property_id (required), +1 more.
Rename a page without changing its body or version metadata.
Args:
page_id: The ID of the page to rename.
title: The new title for the page.
Inputs: title (required), page_id (required).
Update an existing content property on a space.
Args:
space_id: The ID of the space.
property_id: The ID of the property to update.
key: Property key.
value: New property value.
version_number: Current version number of the property (incremented automatically).
Inputs: key (required), value (required), space_id (required), property_id (required), +1 more.
Update a task's status, assignee, or due date.
Args:
task_id: Task ID.
status: New status — 'open' or 'complete'.
assigned_to: Account ID of the new assignee.
due_at: ISO 8601 due date.
Inputs: status (required), task_id (required), due_at, assigned_to.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 62 tools
Several tools are redundant or have overlapping functionality, such as list_pages vs list_pages_in_space and update_page vs update_page_title, which both can modify titles. The large number of similar list/get/create/update/delete tools across pages, blogposts, comments, and properties creates potential confusion despite detailed descriptions.
The vast majority of tools follow a clear verb_noun pattern (e.g., list_pages, create_page, delete_inline_comment), and the naming is consistently snake_case. Minor deviations like 'bulk_user_lookup' (which starts with an adjective rather than a verb) prevent a perfect score, but the overall pattern is predictable.
With 62 tools, the server is extremely large and exceeds the typical well-scoped MCP server by a wide margin. Even for a complex domain like Confluence, this number of tools would overwhelm agents and make tool selection difficult, fitting the 'extreme mismatch' category.
The tool set covers many core Confluence operations including pages, blog posts, spaces, comments, properties, labels, and tasks. However, there are glaring gaps given the size: no search functionality, no attachment upload/update, no task creation/deletion, no space update/delete, and no dedicated blog post comment listing.
An MCP server that provides access to Agility CMS. See https://mcp.agilitycms.com for more details.
A MCP server built for developers enabling Git based project management with project and personal…
Shared project wiki for AI agents: read and write pages, next actions, and activity logs over MCP.
An MCP server that provides an API to LLMs to manage their JumpCloud resources.
