Confluence
Service domainDOCUMENTS
Arcade Optimized
Arcade.dev LLM tools for Confluence
14tools
Confluence Toolkit
The Confluence toolkit lets LLMs interact with Atlassian Confluence via Arcade, enabling agents to read, write, search, and manage pages, spaces, and attachments.
Capabilities
- Page management: Create pages, retrieve single or multiple pages by ID or title, update content, rename pages, and explore hierarchies through direct-child traversal.
- Search: Full-workspace content search with flexible AND/OR logic across all Confluence content, case-insensitive.
- Spaces & clouds: List and inspect spaces by ID or key; discover available Atlassian Cloud instances for the authenticated user.
- Attachments: List workspace attachments and retrieve attachments scoped to a specific page by ID or title.
- User context: Fetch the current user's profile and cloud access via a dedicated identity tool, recommended as a first call in any session.
OAuth
This toolkit uses OAuth 2.0 via the Atlassian provider. Arcade handles the OAuth flow automatically. See the Atlassian auth provider docs for configuration details.
Available tools(14)
14 of 14 tools
Operations
Behavior
| Tool name | Description | Secrets | |
|---|---|---|---|
Create a new page at the root of the given space. | |||
Get attachments for a page by its ID or title.
If a page title is provided, then the first page with an exact matching title will be returned. | |||
Get available Atlassian Clouds. | |||
Retrieve a SINGLE page's content by its ID or title.
If a title is provided, then the first page with an exact matching title will be returned.
IMPORTANT: For retrieving MULTIPLE pages, use `get_pages_by_id` instead
for a massive performance and efficiency boost. If you call this function multiple times
instead of using `get_pages_by_id`, then the universe will explode. | |||
Get the content of MULTIPLE pages by their ID in a single efficient request.
IMPORTANT: Always use this function when you need to retrieve content from more than one page,
rather than making multiple separate calls to get_page, because this function is significantly
more efficient than calling get_page multiple times. | |||
Get the details of a space by its ID or key. | |||
List attachments in a workspace | |||
List one page of direct children for a Confluence content item.
Call this again with a returned child ID to continue exploring that child's children. For
pagination, send only pagination_token and optionally limit. | |||
Get the content of multiple pages by their ID | |||
List all spaces sorted by name in ascending order. | |||
Rename a page by changing its title. | |||
Search for content in Confluence.
The search is performed across all content in the authenticated user's Confluence workspace.
All search terms in Confluence are case insensitive.
You can use the parameters in different ways:
- must_contain_all: For AND logic - content must contain ALL of these
- can_contain_any: For OR logic - content can contain ANY of these
- Combine them: must_contain_all=['banana'] AND can_contain_any=['database', 'guide'] | |||
Update a page's content. | |||
CALL THIS TOOL FIRST to establish user profile context.
Get information about the currently logged-in user and their available Confluence clouds. |
Last updated on